path: root/doc/manual/manual.xmli

<?xml version="1.0"?>
<book xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0">

  <info>
    <title>Nix Package Manager Guide</title>
    <subtitle>Version 3.0</subtitle>

    <author>
      <personname>
        <firstname>Eelco</firstname>
        <surname>Dolstra</surname>
      </personname>
      <contrib>Author</contrib>
    </author>

    <copyright>
      <year>2004-2018</year>
      <holder>Eelco Dolstra</holder>
    </copyright>

  </info>

  <!--
  <preface>
    <title>Preface</title>
    <para>This manual describes how to set up and use the Nix package
    manager.</para>
  </preface>
  -->

  <part xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="chap-introduction" xml:base="introduction/introduction.xml">

<title>Introduction</title>

<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-about-nix">

<title>About Nix</title>

<para>Nix is a <emphasis>purely functional package manager</emphasis>.
This means that it treats packages like values in purely functional
programming languages such as Haskell &#x2014; they are built by functions
that don&#x2019;t have side-effects, and they never change after they have
been built.  Nix stores packages in the <emphasis>Nix
store</emphasis>, usually the directory
<filename>/nix/store</filename>, where each package has its own unique
subdirectory such as

<programlisting>
/nix/store/b6gvzjyb2pg0kjfwrjmg1vfhh54ad73z-firefox-33.1/
</programlisting>

where <literal>b6gvzjyb2pg0&#x2026;</literal> is a unique identifier for the
package that captures all its dependencies (it&#x2019;s a cryptographic hash
of the package&#x2019;s build dependency graph).  This enables many powerful
features.</para>


<simplesect><title>Multiple versions</title>

<para>You can have multiple versions or variants of a package
installed at the same time.  This is especially important when
different applications have dependencies on different versions of the
same package &#x2014; it prevents the &#x201C;DLL hell&#x201D;.  Because of the hashing
scheme, different versions of a package end up in different paths in
the Nix store, so they don&#x2019;t interfere with each other.</para>

<para>An important consequence is that operations like upgrading or
uninstalling an application cannot break other applications, since
these operations never &#x201C;destructively&#x201D; update or delete files that are
used by other packages.</para>

</simplesect>


<simplesect><title>Complete dependencies</title>

<para>Nix helps you make sure that package dependency specifications
are complete.  In general, when you&#x2019;re making a package for a package
management system like RPM, you have to specify for each package what
its dependencies are, but there are no guarantees that this
specification is complete.  If you forget a dependency, then the
package will build and work correctly on <emphasis>your</emphasis>
machine if you have the dependency installed, but not on the end
user's machine if it's not there.</para>

<para>Since Nix on the other hand doesn&#x2019;t install packages in &#x201C;global&#x201D;
locations like <filename>/usr/bin</filename> but in package-specific
directories, the risk of incomplete dependencies is greatly reduced.
This is because tools such as compilers don&#x2019;t search in per-packages
directories such as
<filename>/nix/store/5lbfaxb722zp&#x2026;-openssl-0.9.8d/include</filename>,
so if a package builds correctly on your system, this is because you
specified the dependency explicitly. This takes care of the build-time
dependencies.</para>

<para>Once a package is built, runtime dependencies are found by
scanning binaries for the hash parts of Nix store paths (such as
<literal>r8vvq9kq&#x2026;</literal>).  This sounds risky, but it works
extremely well.</para>

</simplesect>


<simplesect><title>Multi-user support</title>

<para>Nix has multi-user support.  This means that non-privileged
users can securely install software.  Each user can have a different
<emphasis>profile</emphasis>, a set of packages in the Nix store that
appear in the user&#x2019;s <envar>PATH</envar>.  If a user installs a
package that another user has already installed previously, the
package won&#x2019;t be built or downloaded a second time.  At the same time,
it is not possible for one user to inject a Trojan horse into a
package that might be used by another user.</para>

</simplesect>


<simplesect><title>Atomic upgrades and rollbacks</title>

<para>Since package management operations never overwrite packages in
the Nix store but just add new versions in different paths, they are
<emphasis>atomic</emphasis>.  So during a package upgrade, there is no
time window in which the package has some files from the old version
and some files from the new version &#x2014; which would be bad because a
program might well crash if it&#x2019;s started during that period.</para>

<para>And since packages aren&#x2019;t overwritten, the old versions are still
there after an upgrade.  This means that you can <emphasis>roll
back</emphasis> to the old version:</para>

<screen>
$ nix-env --upgrade <replaceable>some-packages</replaceable>
$ nix-env --rollback
</screen>

</simplesect>


<simplesect><title>Garbage collection</title>

<para>When you uninstall a package like this&#x2026;

<screen>
$ nix-env --uninstall firefox
</screen>

the package isn&#x2019;t deleted from the system right away (after all, you
might want to do a rollback, or it might be in the profiles of other
users).  Instead, unused packages can be deleted safely by running the
<emphasis>garbage collector</emphasis>:

<screen>
$ nix-collect-garbage
</screen>

This deletes all packages that aren&#x2019;t in use by any user profile or by
a currently running program.</para>

</simplesect>


<simplesect><title>Functional package language</title>

<para>Packages are built from <emphasis>Nix expressions</emphasis>,
which is a simple functional language.  A Nix expression describes
everything that goes into a package build action (a &#x201C;derivation&#x201D;):
other packages, sources, the build script, environment variables for
the build script, etc.  Nix tries very hard to ensure that Nix
expressions are <emphasis>deterministic</emphasis>: building a Nix
expression twice should yield the same result.</para>

<para>Because it&#x2019;s a functional language, it&#x2019;s easy to support
building variants of a package: turn the Nix expression into a
function and call it any number of times with the appropriate
arguments.  Due to the hashing scheme, variants don&#x2019;t conflict with
each other in the Nix store.</para>

</simplesect>


<simplesect><title>Transparent source/binary deployment</title>

<para>Nix expressions generally describe how to build a package from
source, so an installation action like

<screen>
$ nix-env --install firefox
</screen>

<emphasis>could</emphasis> cause quite a bit of build activity, as not
only Firefox but also all its dependencies (all the way up to the C
library and the compiler) would have to built, at least if they are
not already in the Nix store.  This is a <emphasis>source deployment
model</emphasis>.  For most users, building from source is not very
pleasant as it takes far too long.  However, Nix can automatically
skip building from source and instead use a <emphasis>binary
cache</emphasis>, a web server that provides pre-built binaries. For
instance, when asked to build
<literal>/nix/store/b6gvzjyb2pg0&#x2026;-firefox-33.1</literal> from source,
Nix would first check if the file
<uri>https://cache.nixos.org/b6gvzjyb2pg0&#x2026;.narinfo</uri> exists, and
if so, fetch the pre-built binary referenced from there; otherwise, it
would fall back to building from source.</para>

</simplesect>


<!--
<simplesect><title>Binary patching</title>

<para>In addition to downloading binaries automatically if they’re
available, Nix can download binary deltas that patch an existing
package in the Nix store into a new version.  This speeds up
upgrades.</para>

</simplesect>
-->


<simplesect><title>Nix Packages collection</title>

<para>We provide a large set of Nix expressions containing hundreds of
existing Unix packages, the <emphasis>Nix Packages
collection</emphasis> (Nixpkgs).</para>

</simplesect>


<simplesect><title>Managing build environments</title>

<para>Nix is extremely useful for developers as it makes it easy to
automatically set up the build environment for a package. Given a
Nix expression that describes the dependencies of your package, the
command <command>nix-shell</command> will build or download those
dependencies if they&#x2019;re not already in your Nix store, and then start
a Bash shell in which all necessary environment variables (such as
compiler search paths) are set.</para>

<para>For example, the following command gets all dependencies of the
Pan newsreader, as described by <link xlink:href="https://github.com/NixOS/nixpkgs/blob/master/pkgs/applications/networking/newsreaders/pan/default.nix">its
Nix expression</link>:</para>

<screen>
$ nix-shell '&lt;nixpkgs&gt;' -A pan
</screen>

<para>You&#x2019;re then dropped into a shell where you can edit, build and test
the package:</para>

<screen>
[nix-shell]$ tar xf $src
[nix-shell]$ cd pan-*
[nix-shell]$ ./configure
[nix-shell]$ make
[nix-shell]$ ./pan/gui/pan
</screen>

<!--
<para>Since Nix packages are reproducible and have complete dependency
specifications, Nix makes an excellent basis for <a
href="[%root%]hydra">a continuous build system</a>.</para>
-->

</simplesect>


<simplesect><title>Portability</title>

<para>Nix runs on Linux and macOS.</para>

</simplesect>


<simplesect><title>NixOS</title>

<para>NixOS is a Linux distribution based on Nix.  It uses Nix not
just for package management but also to manage the system
configuration (e.g., to build configuration files in
<filename>/etc</filename>).  This means, among other things, that it
is easy to roll back the entire configuration of the system to an
earlier state.  Also, users can install software without root
privileges.  For more information and downloads, see the <link xlink:href="http://nixos.org/">NixOS homepage</link>.</para>

</simplesect>


<simplesect><title>License</title>

<para>Nix is released under the terms of the <link xlink:href="http://www.gnu.org/licenses/old-licenses/lgpl-2.1.html">GNU
LGPLv2.1 or (at your option) any later version</link>.</para>

</simplesect>


</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="chap-quick-start">

<title>Quick Start</title>

<para>This chapter is for impatient people who don't like reading
documentation.  For more in-depth information you are kindly referred
to subsequent chapters.</para>

<procedure>

<step><para>Install single-user Nix by running the following:

<screen>
$ bash &lt;(curl -L https://nixos.org/nix/install)
</screen>

This will install Nix in <filename>/nix</filename>. The install script
will create <filename>/nix</filename> using <command>sudo</command>,
so make sure you have sufficient rights.  (For other installation
methods, see <xref linkend="chap-installation"/>.)</para></step>

<step><para>See what installable packages are currently available
in the channel:

<screen>
$ nix-env -qa
docbook-xml-4.3
docbook-xml-4.5
firefox-33.0.2
hello-2.9
libxslt-1.1.28
<replaceable>...</replaceable></screen>

</para></step>

<step><para>Install some packages from the channel:

<screen>
$ nix-env -i hello</screen>

This should download pre-built packages; it should not build them
locally (if it does, something went wrong).</para></step>

<step><para>Test that they work:

<screen>
$ which hello
/home/eelco/.nix-profile/bin/hello
$ hello
Hello, world!
</screen>

</para></step>

<step><para>Uninstall a package:

<screen>
$ nix-env -e hello</screen>

</para></step>

<step><para>You can also test a package without installing it:

<screen>
$ nix-shell -p hello
</screen>

This builds or downloads GNU Hello and its dependencies, then drops
you into a Bash shell where the <command>hello</command> command is
present, all without affecting your normal environment:

<screen>
[nix-shell:~]$ hello
Hello, world!

[nix-shell:~]$ exit

$ hello
hello: command not found
</screen>

</para></step>

<step><para>To keep up-to-date with the channel, do:

<screen>
$ nix-channel --update nixpkgs
$ nix-env -u '*'</screen>

The latter command will upgrade each installed package for which there
is a &#x201C;newer&#x201D; version (as determined by comparing the version
numbers).</para></step>

<step><para>If you're unhappy with the result of a
<command>nix-env</command> action (e.g., an upgraded package turned
out not to work properly), you can go back:

<screen>
$ nix-env --rollback</screen>

</para></step>

<step><para>You should periodically run the Nix garbage collector
to get rid of unused packages, since uninstalls or upgrades don't
actually delete them:

<screen>
$ nix-collect-garbage -d</screen>

<!--
The first command deletes old “generations” of your profile (making
rollbacks impossible, but also making the packages in those old
generations available for garbage collection), while the second
command actually deletes them.-->

</para></step>

</procedure>

</chapter>

</part>
  <part xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="chap-installation" xml:base="installation/installation.xml">

<title>Installation</title>

<partintro>
<para>This section describes how to install and configure Nix for first-time use.</para>
</partintro>

<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-supported-platforms">

<title>Supported Platforms</title>

<para>Nix is currently supported on the following platforms:

<itemizedlist>

  <listitem><para>Linux (i686, x86_64, aarch64).</para></listitem>

  <listitem><para>macOS (x86_64).</para></listitem>

  <!--
  <listitem><para>FreeBSD (only tested on Intel).</para></listitem>
  -->

  <!--
  <listitem><para>Windows through <link
  xlink:href="http://www.cygwin.com/">Cygwin</link>.</para>

  <warning><para>On Cygwin, Nix <emphasis>must</emphasis> be installed
  on an NTFS partition.  It will not work correctly on a FAT
  partition.</para></warning>

  </listitem>
  -->

</itemizedlist>

</para>

</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-installing-binary">

<title>Installing a Binary Distribution</title>

<para>
  If you are using Linux or macOS versions up to 10.14 (Mojave), the
  easiest way to install Nix is to run the following command:
</para>

<screen>
  $ sh &lt;(curl -L https://nixos.org/nix/install)
</screen>

<para>
  If you're using macOS 10.15 (Catalina) or newer, consult
  <link linkend="sect-macos-installation">the macOS installation instructions</link>
  before installing.
</para>

<para>
  As of Nix 2.1.0, the Nix installer will always default to creating a
  single-user installation, however opting in to the multi-user
  installation is highly recommended.
  <!-- TODO: this explains *neither* why the default version is
  single-user, nor why we'd recommend multi-user over the default.
  True prospective users don't have much basis for evaluating this.
  What's it to me? Who should pick which? Why? What if I pick wrong?
  -->
</para>

<section xml:id="sect-single-user-installation">
  <title>Single User Installation</title>

  <para>
    To explicitly select a single-user installation on your system:

    <screen>
  sh &lt;(curl -L https://nixos.org/nix/install) --no-daemon
</screen>
  </para>

<para>
This will perform a single-user installation of Nix, meaning that
<filename>/nix</filename> is owned by the invoking user.  You should
run this under your usual user account, <emphasis>not</emphasis> as
root.  The script will invoke <command>sudo</command> to create
<filename>/nix</filename> if it doesn&#x2019;t already exist.  If you don&#x2019;t
have <command>sudo</command>, you should manually create
<filename>/nix</filename> first as root, e.g.:

<screen>
$ mkdir /nix
$ chown alice /nix
</screen>

The install script will modify the first writable file from amongst
<filename>.bash_profile</filename>, <filename>.bash_login</filename>
and <filename>.profile</filename> to source
<filename>~/.nix-profile/etc/profile.d/nix.sh</filename>. You can set
the <envar>NIX_INSTALLER_NO_MODIFY_PROFILE</envar> environment
variable before executing the install script to disable this
behaviour.
</para>


<para>You can uninstall Nix simply by running:

<screen>
$ rm -rf /nix
</screen>

</para>
</section>

<section xml:id="sect-multi-user-installation">
  <title>Multi User Installation</title>
  <para>
    The multi-user Nix installation creates system users, and a system
    service for the Nix daemon.
  </para>

  <itemizedlist>
    <title>Supported Systems</title>

    <listitem>
      <para>Linux running systemd, with SELinux disabled</para>
    </listitem>
    <listitem><para>macOS</para></listitem>
  </itemizedlist>

  <para>
    You can instruct the installer to perform a multi-user
    installation on your system:
  </para>

  <screen>sh &lt;(curl -L https://nixos.org/nix/install) --daemon</screen>

  <para>
    The multi-user installation of Nix will create build users between
    the user IDs 30001 and 30032, and a group with the group ID 30000.

    You should run this under your usual user account,
    <emphasis>not</emphasis> as root. The script will invoke
    <command>sudo</command> as needed.
  </para>

  <note><para>
    If you need Nix to use a different group ID or user ID set, you
    will have to download the tarball manually and <link linkend="sect-nix-install-binary-tarball">edit the install
    script</link>.
  </para></note>

  <para>
    The installer will modify <filename>/etc/bashrc</filename>, and
    <filename>/etc/zshrc</filename> if they exist. The installer will
    first back up these files with a
    <literal>.backup-before-nix</literal> extension. The installer
    will also create <filename>/etc/profile.d/nix.sh</filename>.
  </para>

  <para>You can uninstall Nix with the following commands:

<screen>
sudo rm -rf /etc/profile/nix.sh /etc/nix /nix ~root/.nix-profile ~root/.nix-defexpr ~root/.nix-channels ~/.nix-profile ~/.nix-defexpr ~/.nix-channels

# If you are on Linux with systemd, you will need to run:
sudo systemctl stop nix-daemon.socket
sudo systemctl stop nix-daemon.service
sudo systemctl disable nix-daemon.socket
sudo systemctl disable nix-daemon.service
sudo systemctl daemon-reload

# If you are on macOS, you will need to run:
sudo launchctl unload /Library/LaunchDaemons/org.nixos.nix-daemon.plist
sudo rm /Library/LaunchDaemons/org.nixos.nix-daemon.plist
</screen>

    There may also be references to Nix in
    <filename>/etc/profile</filename>,
    <filename>/etc/bashrc</filename>, and
    <filename>/etc/zshrc</filename> which you may remove.
  </para>

</section>

<section xml:id="sect-macos-installation">
  <title>macOS Installation</title>

  <para>
    Starting with macOS 10.15 (Catalina), the root filesystem is read-only.
    This means <filename>/nix</filename> can no longer live on your system
    volume, and that you'll need a workaround to install Nix.
  </para>

  <para>
    The recommended approach, which creates an unencrypted APFS volume
    for your Nix store and a "synthetic" empty directory to mount it
    over at <filename>/nix</filename>, is least likely to impair Nix
    or your system.
  </para>

  <note><para>
    With all separate-volume approaches, it's possible something on
    your system (particularly daemons/services and restored apps) may
    need access to your Nix store before the volume is mounted. Adding
    additional encryption makes this more likely.
  </para></note>

  <para>
    If you're using a recent Mac with a
    <link xlink:href="https://www.apple.com/euro/mac/shared/docs/Apple_T2_Security_Chip_Overview.pdf">T2 chip</link>,
    your drive will still be encrypted at rest (in which case "unencrypted"
    is a bit of a misnomer). To use this approach, just install Nix with:
  </para>

  <screen>$ sh &lt;(curl -L https://nixos.org/nix/install) --darwin-use-unencrypted-nix-store-volume</screen>

  <para>
    If you don't like the sound of this, you'll want to weigh the
    other approaches and tradeoffs detailed in this section.
  </para>

  <note>
    <title>Eventual solutions?</title>
    <para>
      All of the known workarounds have drawbacks, but we hope
      better solutions will be available in the future. Some that
      we have our eye on are:
    </para>
    <orderedlist>
      <listitem>
        <para>
          A true firmlink would enable the Nix store to live on the
          primary data volume without the build problems caused by
          the symlink approach. End users cannot currently
          create true firmlinks.
        </para>
      </listitem>
      <listitem>
        <para>
          If the Nix store volume shared FileVault encryption
          with the primary data volume (probably by using the same
          volume group and role), FileVault encryption could be
          easily supported by the installer without requiring
          manual setup by each user.
        </para>
      </listitem>
    </orderedlist>
  </note>

  <section xml:id="sect-macos-installation-change-store-prefix">
    <title>Change the Nix store path prefix</title>
    <para>
      Changing the default prefix for the Nix store is a simple
      approach which enables you to leave it on your root volume,
      where it can take full advantage of FileVault encryption if
      enabled. Unfortunately, this approach also opts your device out
      of some benefits that are enabled by using the same prefix
      across systems:

      <itemizedlist>
        <listitem>
          <para>
            Your system won't be able to take advantage of the binary
            cache (unless someone is able to stand up and support
            duplicate caching infrastructure), which means you'll
            spend more time waiting for builds.
          </para>
        </listitem>
        <listitem>
          <para>
            It's harder to build and deploy packages to Linux systems.
          </para>
        </listitem>
        <!-- TODO: may be more here -->
      </itemizedlist>

      <!-- TODO: Yes, but how?! -->

      It would also possible (and often requested) to just apply this
      change ecosystem-wide, but it's an intrusive process that has
      side effects we want to avoid for now.
      <!-- magnificent hand-wavy gesture -->
    </para>
    <para>
    </para>
  </section>

  <section xml:id="sect-macos-installation-encrypted-volume">
    <title>Use a separate encrypted volume</title>
    <para>
      If you like, you can also add encryption to the recommended
      approach taken by the installer. You can do this by pre-creating
      an encrypted volume before you run the installer--or you can
      run the installer and encrypt the volume it creates later.
      <!-- TODO: see later note about whether this needs both add-encryption and from-scratch directions -->
    </para>
    <para>
      In either case, adding encryption to a second volume isn't quite
      as simple as enabling FileVault for your boot volume. Before you
      dive in, there are a few things to weigh:
    </para>
    <orderedlist>
      <listitem>
        <para>
          The additional volume won't be encrypted with your existing
          FileVault key, so you'll need another mechanism to decrypt
          the volume.
        </para>
      </listitem>
      <listitem>
        <para>
          You can store the password in Keychain to automatically
          decrypt the volume on boot--but it'll have to wait on Keychain
          and may not mount before your GUI apps restore. If any of
          your launchd agents or apps depend on Nix-installed software
          (for example, if you use a Nix-installed login shell), the
          restore may fail or break.
        </para>
        <para>
          On a case-by-case basis, you may be able to work around this
          problem by using <command>wait4path</command> to block
          execution until your executable is available.
        </para>
        <para>
          It's also possible to decrypt and mount the volume earlier
          with a login hook--but this mechanism appears to be
          deprecated and its future is unclear.
        </para>
      </listitem>
      <listitem>
        <para>
          You can hard-code the password in the clear, so that your
          store volume can be decrypted before Keychain is available.
        </para>
      </listitem>
    </orderedlist>
    <para>
      If you are comfortable navigating these tradeoffs, you can encrypt the volume with
      something along the lines of:
      <!-- TODO:
      I don't know if this also needs from-scratch instructions?
      can we just recommend use-the-installer-and-then-encrypt?
      -->
    </para>
    <!--
    TODO: it looks like this option can be encryptVolume|encrypt|enableFileVault

    It may be more clear to use encryptVolume, here? FileVault seems
    heavily associated with the boot-volume behavior; I worry
    a little that it can mislead here, especially as it gets
    copied around minus doc context...?
    -->
    <screen>alice$ diskutil apfs enableFileVault /nix -user disk</screen>

    <!-- TODO: and then go into detail on the mount/decrypt approaches? -->
  </section>

  <section xml:id="sect-macos-installation-symlink">
    <!--
    Maybe a good razor is: if we'd hate having to support someone who
    installed Nix this way, it shouldn't even be detailed?
    -->
    <title>Symlink the Nix store to a custom location</title>
    <para>
      Another simple approach is using <filename>/etc/synthetic.conf</filename>
      to symlink the Nix store to the data volume. This option also
      enables your store to share any configured FileVault encryption.
      Unfortunately, builds that resolve the symlink may leak the
      canonical path or even fail.
    </para>
    <para>
      Because of these downsides, we can't recommend this approach.
    </para>
    <!-- Leaving out instructions for this one. -->
  </section>

  <section xml:id="sect-macos-installation-recommended-notes">
    <title>Notes on the recommended approach</title>
    <para>
      This section goes into a little more detail on the recommended
      approach. You don't need to understand it to run the installer,
      but it can serve as a helpful reference if you run into trouble.
    </para>
    <orderedlist>
      <listitem>
        <para>
          In order to compose user-writable locations into the new
          read-only system root, Apple introduced a new concept called
          <literal>firmlinks</literal>, which it describes as a
          "bi-directional wormhole" between two filesystems. You can
          see the current firmlinks in <filename>/usr/share/firmlinks</filename>.
          Unfortunately, firmlinks aren't (currently?) user-configurable.
        </para>

        <para>
          For special cases like NFS mount points or package manager roots,
          <link xlink:href="https://developer.apple.com/library/archive/documentation/System/Conceptual/ManPages_iPhoneOS/man5/synthetic.conf.5.html">synthetic.conf(5)</link>
          supports limited user-controlled file-creation (of symlinks,
          and synthetic empty directories) at <filename>/</filename>.
          To create a synthetic empty directory for mounting at <filename>/nix</filename>,
          add the following line to <filename>/etc/synthetic.conf</filename>
          (create it if necessary):
        </para>

        <screen>nix</screen>
      </listitem>

      <listitem>
        <para>
          This configuration is applied at boot time, but you can use
          <command>apfs.util</command> to trigger creation (not deletion)
          of new entries without a reboot:
        </para>

        <screen>alice$ /System/Library/Filesystems/apfs.fs/Contents/Resources/apfs.util -B</screen>
      </listitem>

      <listitem>
        <para>
          Create the new APFS volume with diskutil:
        </para>

        <screen>alice$ sudo diskutil apfs addVolume diskX APFS 'Nix Store' -mountpoint /nix</screen>
      </listitem>

      <listitem>
        <para>
          Using <command>vifs</command>, add the new mount to
          <filename>/etc/fstab</filename>. If it doesn't already have
          other entries, it should look something like:
        </para>

<screen>
#
# Warning - this file should only be modified with vifs(8)
#
# Failure to do so is unsupported and may be destructive.
#
LABEL=Nix\040Store /nix apfs rw,nobrowse
</screen>

        <para>
          The nobrowse setting will keep Spotlight from indexing this
          volume, and keep it from showing up on your desktop.
        </para>
      </listitem>
    </orderedlist>
  </section>

</section>

<section xml:id="sect-nix-install-pinned-version-url">
  <title>Installing a pinned Nix version from a URL</title>

  <para>
    NixOS.org hosts version-specific installation URLs for all Nix
    versions since 1.11.16, at
    <literal>https://releases.nixos.org/nix/nix-<replaceable>version</replaceable>/install</literal>.
  </para>

  <para>
    These install scripts can be used the same as the main
  NixOS.org installation script:

  <screen>
  sh &lt;(curl -L https://nixos.org/nix/install)
</screen>
  </para>

  <para>
    In the same directory of the install script are sha256 sums, and
    gpg signature files.
  </para>
</section>

<section xml:id="sect-nix-install-binary-tarball">
  <title>Installing from a binary tarball</title>

  <para>
    You can also download a binary tarball that contains Nix and all
    its dependencies.  (This is what the install script at
    <uri>https://nixos.org/nix/install</uri> does automatically.)  You
    should unpack it somewhere (e.g. in <filename>/tmp</filename>),
    and then run the script named <command>install</command> inside
    the binary tarball:


<screen>
alice$ cd /tmp
alice$ tar xfj nix-1.8-x86_64-darwin.tar.bz2
alice$ cd nix-1.8-x86_64-darwin
alice$ ./install
</screen>
  </para>

  <para>
    If you need to edit the multi-user installation script to use
    different group ID or a different user ID range, modify the
    variables set in the file named
    <filename>install-multi-user</filename>.
  </para>
</section>
</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-installing-source">

<title>Installing Nix from Source</title>

<para>If no binary package is available, you can download and compile
a source distribution.</para>

<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-prerequisites-source">

<title>Prerequisites</title>

<itemizedlist>

  <listitem><para>GNU Autoconf
  (<link xlink:href="https://www.gnu.org/software/autoconf/"/>)
  and the autoconf-archive macro collection
  (<link xlink:href="https://www.gnu.org/software/autoconf-archive/"/>).
  These are only needed to run the bootstrap script, and are not necessary
  if your source distribution came with a pre-built
  <literal>./configure</literal> script.</para></listitem>

  <listitem><para>GNU Make.</para></listitem>
  
  <listitem><para>Bash Shell. The <literal>./configure</literal> script
  relies on bashisms, so Bash is required.</para></listitem>

  <listitem><para>A version of GCC or Clang that supports C++17.</para></listitem>

  <listitem><para><command>pkg-config</command> to locate
  dependencies.  If your distribution does not provide it, you can get
  it from <link xlink:href="http://www.freedesktop.org/wiki/Software/pkg-config"/>.</para></listitem>

  <listitem><para>The OpenSSL library to calculate cryptographic hashes.
  If your distribution does not provide it, you can get it from <link xlink:href="https://www.openssl.org"/>.</para></listitem>

  <listitem><para>The <literal>libbrotlienc</literal> and
  <literal>libbrotlidec</literal> libraries to provide implementation
  of the Brotli compression algorithm. They are available for download
  from the official repository <link xlink:href="https://github.com/google/brotli"/>.</para></listitem>

  <listitem><para>The bzip2 compressor program and the
  <literal>libbz2</literal> library.  Thus you must have bzip2
  installed, including development headers and libraries.  If your
  distribution does not provide these, you can obtain bzip2 from <link xlink:href="https://web.archive.org/web/20180624184756/http://www.bzip.org/"/>.</para></listitem>

  <listitem><para><literal>liblzma</literal>, which is provided by
  XZ Utils. If your distribution does not provide this, you can
  get it from <link xlink:href="https://tukaani.org/xz/"/>.</para></listitem>
  
  <listitem><para>cURL and its library. If your distribution does not
  provide it, you can get it from <link xlink:href="https://curl.haxx.se/"/>.</para></listitem>
      
  <listitem><para>The SQLite embedded database library, version 3.6.19
  or higher.  If your distribution does not provide it, please install
  it from <link xlink:href="http://www.sqlite.org/"/>.</para></listitem>

  <listitem><para>The <link xlink:href="http://www.hboehm.info/gc/">Boehm
  garbage collector</link> to reduce the evaluator&#x2019;s memory
  consumption (optional).  To enable it, install
  <literal>pkgconfig</literal> and the Boehm garbage collector, and
  pass the flag <option>--enable-gc</option> to
  <command>configure</command>.</para></listitem>

  <listitem><para>The <literal>boost</literal> library of version
  1.66.0 or higher. It can be obtained from the official web site
  <link xlink:href="https://www.boost.org/"/>.</para></listitem>

  <listitem><para>The <literal>editline</literal> library of version
  1.14.0 or higher. It can be obtained from the its repository
  <link xlink:href="https://github.com/troglobit/editline"/>.</para></listitem>

  <listitem><para>The <command>xmllint</command> and
  <command>xsltproc</command> programs to build this manual and the
  man-pages.  These are part of the <literal>libxml2</literal> and
  <literal>libxslt</literal> packages, respectively.  You also need
  the <link xlink:href="http://docbook.sourceforge.net/projects/xsl/">DocBook
  XSL stylesheets</link> and optionally the <link xlink:href="http://www.docbook.org/schemas/5x"> DocBook 5.0 RELAX NG
  schemas</link>.  Note that these are only required if you modify the
  manual sources or when you are building from the Git
  repository.</para></listitem>

  <listitem><para>Recent versions of Bison and Flex to build the
  parser.  (This is because Nix needs GLR support in Bison and
  reentrancy support in Flex.)  For Bison, you need version 2.6, which
  can be obtained from the <link xlink:href="ftp://alpha.gnu.org/pub/gnu/bison">GNU FTP
  server</link>.  For Flex, you need version 2.5.35, which is
  available on <link xlink:href="http://lex.sourceforge.net/">SourceForge</link>.
  Slightly older versions may also work, but ancient versions like the
  ubiquitous 2.5.4a won't.  Note that these are only required if you
  modify the parser or when you are building from the Git
  repository.</para></listitem>

  <listitem><para>The <literal>libseccomp</literal> is used to provide
  syscall filtering on Linux. This is an optional dependency and can
  be disabled passing a <option>--disable-seccomp-sandboxing</option>
  option to the <command>configure</command> script (Not recommended
  unless your system doesn't support
  <literal>libseccomp</literal>). To get the library, visit <link xlink:href="https://github.com/seccomp/libseccomp"/>.</para></listitem>

</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-obtaining-source">

<title>Obtaining a Source Distribution</title>

<para>The source tarball of the most recent stable release can be
downloaded from the <link xlink:href="http://nixos.org/nix/download.html">Nix homepage</link>.
You can also grab the <link xlink:href="http://hydra.nixos.org/job/nix/master/release/latest-finished#tabs-constituents">most
recent development release</link>.</para>

<para>Alternatively, the most recent sources of Nix can be obtained
from its <link xlink:href="https://github.com/NixOS/nix">Git
repository</link>.  For example, the following command will check out
the latest revision into a directory called
<filename>nix</filename>:</para>

<screen>
$ git clone https://github.com/NixOS/nix</screen>

<para>Likewise, specific releases can be obtained from the <link xlink:href="https://github.com/NixOS/nix/tags">tags</link> of the
repository.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-building-source">

<title>Building Nix from Source</title>

<para>After unpacking or checking out the Nix sources, issue the
following commands:

<screen>
$ ./configure <replaceable>options...</replaceable>
$ make
$ make install</screen>

Nix requires GNU Make so you may need to invoke
<command>gmake</command> instead.</para>

<para>When building from the Git repository, these should be preceded
by the command:

<screen>
$ ./bootstrap.sh</screen>

</para>

<para>The installation path can be specified by passing the
<option>--prefix=<replaceable>prefix</replaceable></option> to
<command>configure</command>.  The default installation directory is
<filename>/usr/local</filename>.  You can change this to any location
you like.  You must have write permission to the
<replaceable>prefix</replaceable> path.</para>

<para>Nix keeps its <emphasis>store</emphasis> (the place where
packages are stored) in <filename>/nix/store</filename> by default.
This can be changed using
<option>--with-store-dir=<replaceable>path</replaceable></option>.</para>

<warning><para>It is best <emphasis>not</emphasis> to change the Nix
store from its default, since doing so makes it impossible to use
pre-built binaries from the standard Nixpkgs channels &#x2014; that is, all
packages will need to be built from source.</para></warning>

<para>Nix keeps state (such as its database and log files) in
<filename>/nix/var</filename> by default.  This can be changed using
<option>--localstatedir=<replaceable>path</replaceable></option>.</para>

</section>

</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-nix-security">

<title>Security</title>

<para>Nix has two basic security models.  First, it can be used in
&#x201C;single-user mode&#x201D;, which is similar to what most other package
management tools do: there is a single user (typically <systemitem class="username">root</systemitem>) who performs all package
management operations.  All other users can then use the installed
packages, but they cannot perform package management operations
themselves.</para>

<para>Alternatively, you can configure Nix in &#x201C;multi-user mode&#x201D;.  In
this model, all users can perform package management operations &#x2014; for
instance, every user can install software without requiring root
privileges.  Nix ensures that this is secure.  For instance, it&#x2019;s not
possible for one user to overwrite a package used by another user with
a Trojan horse.</para>

<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-single-user">

<title>Single-User Mode</title>

<para>In single-user mode, all Nix operations that access the database
in <filename><replaceable>prefix</replaceable>/var/nix/db</filename>
or modify the Nix store in
<filename><replaceable>prefix</replaceable>/store</filename> must be
performed under the user ID that owns those directories.  This is
typically <systemitem class="username">root</systemitem>.  (If you
install from RPM packages, that&#x2019;s in fact the default ownership.)
However, on single-user machines, it is often convenient to
<command>chown</command> those directories to your normal user account
so that you don&#x2019;t have to <command>su</command> to <systemitem class="username">root</systemitem> all the time.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-multi-user">

<title>Multi-User Mode</title>

<para>To allow a Nix store to be shared safely among multiple users,
it is important that users are not able to run builders that modify
the Nix store or database in arbitrary ways, or that interfere with
builds started by other users.  If they could do so, they could
install a Trojan horse in some package and compromise the accounts of
other users.</para>

<para>To prevent this, the Nix store and database are owned by some
privileged user (usually <literal>root</literal>) and builders are
executed under special user accounts (usually named
<literal>nixbld1</literal>, <literal>nixbld2</literal>, etc.).  When a
unprivileged user runs a Nix command, actions that operate on the Nix
store (such as builds) are forwarded to a <emphasis>Nix
daemon</emphasis> running under the owner of the Nix store/database
that performs the operation.</para>

<note><para>Multi-user mode has one important limitation: only
<systemitem class="username">root</systemitem> and a set of trusted
users specified in <filename>nix.conf</filename> can specify arbitrary
binary caches. So while unprivileged users may install packages from
arbitrary Nix expressions, they may not get pre-built
binaries.</para></note>


<simplesect>

<title>Setting up the build users</title>

<para>The <emphasis>build users</emphasis> are the special UIDs under
which builds are performed.  They should all be members of the
<emphasis>build users group</emphasis> <literal>nixbld</literal>.
This group should have no other members.  The build users should not
be members of any other group. On Linux, you can create the group and
users as follows:

<screen>
$ groupadd -r nixbld
$ for n in $(seq 1 10); do useradd -c "Nix build user $n" \
    -d /var/empty -g nixbld -G nixbld -M -N -r -s "$(which nologin)" \
    nixbld$n; done
</screen>

This creates 10 build users. There can never be more concurrent builds
than the number of build users, so you may want to increase this if
you expect to do many builds at the same time.</para>

</simplesect>


<simplesect>

<title>Running the daemon</title>

<para>The <link linkend="sec-nix-daemon">Nix daemon</link> should be
started as follows (as <literal>root</literal>):

<screen>
$ nix-daemon</screen>

You&#x2019;ll want to put that line somewhere in your system&#x2019;s boot
scripts.</para>

<para>To let unprivileged users use the daemon, they should set the
<link linkend="envar-remote"><envar>NIX_REMOTE</envar> environment
variable</link> to <literal>daemon</literal>.  So you should put a
line like

<programlisting>
export NIX_REMOTE=daemon</programlisting>

into the users&#x2019; login scripts.</para>

</simplesect>


<simplesect>

<title>Restricting access</title>

<para>To limit which users can perform Nix operations, you can use the
permissions on the directory
<filename>/nix/var/nix/daemon-socket</filename>.  For instance, if you
want to restrict the use of Nix to the members of a group called
<literal>nix-users</literal>, do

<screen>
$ chgrp nix-users /nix/var/nix/daemon-socket
$ chmod ug=rwx,o= /nix/var/nix/daemon-socket
</screen>

This way, users who are not in the <literal>nix-users</literal> group
cannot connect to the Unix domain socket
<filename>/nix/var/nix/daemon-socket/socket</filename>, so they cannot
perform Nix operations.</para>

</simplesect>


</section>

</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-env-variables">

<title>Environment Variables</title>

<para>To use Nix, some environment variables should be set.  In
particular, <envar>PATH</envar> should contain the directories
<filename><replaceable>prefix</replaceable>/bin</filename> and
<filename>~/.nix-profile/bin</filename>.  The first directory contains
the Nix tools themselves, while <filename>~/.nix-profile</filename> is
a symbolic link to the current <emphasis>user environment</emphasis>
(an automatically generated package consisting of symlinks to
installed packages).  The simplest way to set the required environment
variables is to include the file
<filename><replaceable>prefix</replaceable>/etc/profile.d/nix.sh</filename>
in your <filename>~/.profile</filename> (or similar), like this:</para>

<screen>
source <replaceable>prefix</replaceable>/etc/profile.d/nix.sh</screen>

<section xml:id="sec-nix-ssl-cert-file">

<title><envar>NIX_SSL_CERT_FILE</envar></title>

<para>If you need to specify a custom certificate bundle to account
for an HTTPS-intercepting man in the middle proxy, you must specify
the path to the certificate bundle in the environment variable
<envar>NIX_SSL_CERT_FILE</envar>.</para>


<para>If you don't specify a <envar>NIX_SSL_CERT_FILE</envar>
manually, Nix will install and use its own certificate
bundle.</para>

<procedure>
  <step><para>Set the environment variable and install Nix</para>
    <screen>
$ export NIX_SSL_CERT_FILE=/etc/ssl/my-certificate-bundle.crt
$ sh &lt;(curl -L https://nixos.org/nix/install)
</screen></step>

  <step><para>In the shell profile and rc files (for example,
  <filename>/etc/bashrc</filename>, <filename>/etc/zshrc</filename>),
  add the following line:</para>
<programlisting>
export NIX_SSL_CERT_FILE=/etc/ssl/my-certificate-bundle.crt
</programlisting>
</step>
</procedure>

<note><para>You must not add the export and then do the install, as
the Nix installer will detect the presense of Nix configuration, and
abort.</para></note>

<section xml:id="sec-nix-ssl-cert-file-with-nix-daemon-and-macos">
<title><envar>NIX_SSL_CERT_FILE</envar> with macOS and the Nix daemon</title>

<para>On macOS you must specify the environment variable for the Nix
daemon service, then restart it:</para>

<screen>
$ sudo launchctl setenv NIX_SSL_CERT_FILE /etc/ssl/my-certificate-bundle.crt
$ sudo launchctl kickstart -k system/org.nixos.nix-daemon
</screen>
</section>

<section xml:id="sec-installer-proxy-settings">

<title>Proxy Environment Variables</title>

<para>The Nix installer has special handling for these proxy-related
environment variables:
<varname>http_proxy</varname>, <varname>https_proxy</varname>,
<varname>ftp_proxy</varname>, <varname>no_proxy</varname>,
<varname>HTTP_PROXY</varname>, <varname>HTTPS_PROXY</varname>,
<varname>FTP_PROXY</varname>, <varname>NO_PROXY</varname>.
</para>
<para>If any of these variables are set when running the Nix installer,
then the installer will create an override file at
<filename>/etc/systemd/system/nix-daemon.service.d/override.conf</filename>
so <command>nix-daemon</command> will use them.
</para>
</section>

</section>
</chapter>

<!-- TODO: should be updated
<section><title>Upgrading Nix through Nix</title>

<para>You can install the latest stable version of Nix through Nix
itself by subscribing to the channel <link
xlink:href="http://nixos.org/releases/nix/channels/nix-stable" />,
or the latest unstable version by subscribing to the channel <link
xlink:href="http://nixos.org/releases/nix/channels/nix-unstable" />.
You can also do a <link linkend="sec-one-click">one-click
installation</link> by clicking on the package links at <link
xlink:href="http://nixos.org/releases/full-index-nix.html" />.</para>

</section>
-->

</part>
  <chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-upgrading-nix" xml:base="installation/upgrading.xml">

  <title>Upgrading Nix</title>

  <para>
    Multi-user Nix users on macOS can upgrade Nix by running:
    <command>sudo -i sh -c 'nix-channel --update &amp;&amp;
    nix-env -iA nixpkgs.nix &amp;&amp;
    launchctl remove org.nixos.nix-daemon &amp;&amp;
    launchctl load /Library/LaunchDaemons/org.nixos.nix-daemon.plist'</command>
  </para>


  <para>
    Single-user installations of Nix should run this:
    <command>nix-channel --update; nix-env -iA nixpkgs.nix nixpkgs.cacert</command>
  </para>
      
  <para>
    Multi-user Nix users on Linux should run this with sudo:
    <command>nix-channel --update; nix-env -iA nixpkgs.nix nixpkgs.cacert; systemctl daemon-reload; systemctl restart nix-daemon</command>
  </para>
</chapter>
  <part xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="chap-package-management" xml:base="packages/package-management.xml">

<title>Package Management</title>

<partintro>
<para>This chapter discusses how to do package management with Nix,
i.e., how to obtain, install, upgrade, and erase packages.  This is
the &#x201C;user&#x2019;s&#x201D; perspective of the Nix system &#x2014; people
who want to <emphasis>create</emphasis> packages should consult
<xref linkend="chap-writing-nix-expressions"/>.</para>
</partintro>

<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-basic-package-mgmt">

<title>Basic Package Management</title>

<para>The main command for package management is <link linkend="sec-nix-env"><command>nix-env</command></link>.  You can use
it to install, upgrade, and erase packages, and to query what
packages are installed or are available for installation.</para>

<para>In Nix, different users can have different &#x201C;views&#x201D;
on the set of installed applications.  That is, there might be lots of
applications present on the system (possibly in many different
versions), but users can have a specific selection of those active &#x2014;
where &#x201C;active&#x201D; just means that it appears in a directory
in the user&#x2019;s <envar>PATH</envar>.  Such a view on the set of
installed applications is called a <emphasis>user
environment</emphasis>, which is just a directory tree consisting of
symlinks to the files of the active applications.  </para>

<para>Components are installed from a set of <emphasis>Nix
expressions</emphasis> that tell Nix how to build those packages,
including, if necessary, their dependencies.  There is a collection of
Nix expressions called the Nixpkgs package collection that contains
packages ranging from basic development stuff such as GCC and Glibc,
to end-user applications like Mozilla Firefox.  (Nix is however not
tied to the Nixpkgs package collection; you could write your own Nix
expressions based on Nixpkgs, or completely new ones.)</para>

<para>You can manually download the latest version of Nixpkgs from
<link xlink:href="http://nixos.org/nixpkgs/download.html"/>. However,
it&#x2019;s much more convenient to use the Nixpkgs
<emphasis>channel</emphasis>, since it makes it easy to stay up to
date with new versions of Nixpkgs. (Channels are described in more
detail in <xref linkend="sec-channels"/>.) Nixpkgs is automatically
added to your list of &#x201C;subscribed&#x201D; channels when you install
Nix. If this is not the case for some reason, you can add it as
follows:

<screen>
$ nix-channel --add https://nixos.org/channels/nixpkgs-unstable
$ nix-channel --update
</screen>

</para>

<note><para>On NixOS, you&#x2019;re automatically subscribed to a NixOS
channel corresponding to your NixOS major release
(e.g. <uri>http://nixos.org/channels/nixos-14.12</uri>). A NixOS
channel is identical to the Nixpkgs channel, except that it contains
only Linux binaries and is updated only if a set of regression tests
succeed.</para></note>

<para>You can view the set of available packages in Nixpkgs:

<screen>
$ nix-env -qa
aterm-2.2
bash-3.0
binutils-2.15
bison-1.875d
blackdown-1.4.2
bzip2-1.0.2
&#x2026;</screen>

The flag <option>-q</option> specifies a query operation, and
<option>-a</option> means that you want to show the &#x201C;available&#x201D; (i.e.,
installable) packages, as opposed to the installed packages. If you
downloaded Nixpkgs yourself, or if you checked it out from GitHub,
then you need to pass the path to your Nixpkgs tree using the
<option>-f</option> flag:

<screen>
$ nix-env -qaf <replaceable>/path/to/nixpkgs</replaceable>
</screen>

where <replaceable>/path/to/nixpkgs</replaceable> is where you&#x2019;ve
unpacked or checked out Nixpkgs.</para>

<para>You can select specific packages by name:

<screen>
$ nix-env -qa firefox
firefox-34.0.5
firefox-with-plugins-34.0.5
</screen>

and using regular expressions:

<screen>
$ nix-env -qa 'firefox.*'
</screen>

</para>

<para>It is also possible to see the <emphasis>status</emphasis> of
available packages, i.e., whether they are installed into the user
environment and/or present in the system:

<screen>
$ nix-env -qas
&#x2026;
-PS bash-3.0
--S binutils-2.15
IPS bison-1.875d
&#x2026;</screen>

The first character (<literal>I</literal>) indicates whether the
package is installed in your current user environment.  The second
(<literal>P</literal>) indicates whether it is present on your system
(in which case installing it into your user environment would be a
very quick operation).  The last one (<literal>S</literal>) indicates
whether there is a so-called <emphasis>substitute</emphasis> for the
package, which is Nix&#x2019;s mechanism for doing binary deployment.  It
just means that Nix knows that it can fetch a pre-built package from
somewhere (typically a network server) instead of building it
locally.</para>

<para>You can install a package using <literal>nix-env -i</literal>.
For instance,

<screen>
$ nix-env -i subversion</screen>

will install the package called <literal>subversion</literal> (which
is, of course, the <link xlink:href="http://subversion.tigris.org/">Subversion version
management system</link>).</para>

<note><para>When you ask Nix to install a package, it will first try
to get it in pre-compiled form from a <emphasis>binary
cache</emphasis>. By default, Nix will use the binary cache
<uri>https://cache.nixos.org</uri>; it contains binaries for most
packages in Nixpkgs. Only if no binary is available in the binary
cache, Nix will build the package from source. So if <literal>nix-env
-i subversion</literal> results in Nix building stuff from source,
then either the package is not built for your platform by the Nixpkgs
build servers, or your version of Nixpkgs is too old or too new. For
instance, if you have a very recent checkout of Nixpkgs, then the
Nixpkgs build servers may not have had a chance to build everything
and upload the resulting binaries to
<uri>https://cache.nixos.org</uri>. The Nixpkgs channel is only
updated after all binaries have been uploaded to the cache, so if you
stick to the Nixpkgs channel (rather than using a Git checkout of the
Nixpkgs tree), you will get binaries for most packages.</para></note>

<para>Naturally, packages can also be uninstalled:

<screen>
$ nix-env -e subversion</screen>

</para>

<para>Upgrading to a new version is just as easy.  If you have a new
release of Nix Packages, you can do:

<screen>
$ nix-env -u subversion</screen>

This will <emphasis>only</emphasis> upgrade Subversion if there is a
&#x201C;newer&#x201D; version in the new set of Nix expressions, as
defined by some pretty arbitrary rules regarding ordering of version
numbers (which generally do what you&#x2019;d expect of them).  To just
unconditionally replace Subversion with whatever version is in the Nix
expressions, use <parameter>-i</parameter> instead of
<parameter>-u</parameter>; <parameter>-i</parameter> will remove
whatever version is already installed.</para>

<para>You can also upgrade all packages for which there are newer
versions:

<screen>
$ nix-env -u</screen>

</para>

<para>Sometimes it&#x2019;s useful to be able to ask what
<command>nix-env</command> would do, without actually doing it.  For
instance, to find out what packages would be upgraded by
<literal>nix-env -u</literal>, you can do

<screen>
$ nix-env -u --dry-run
(dry run; not doing anything)
upgrading `libxslt-1.1.0' to `libxslt-1.1.10'
upgrading `graphviz-1.10' to `graphviz-1.12'
upgrading `coreutils-5.0' to `coreutils-5.2.1'</screen>

</para>

</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-profiles">

<title>Profiles</title>

<para>Profiles and user environments are Nix&#x2019;s mechanism for
implementing the ability to allow different users to have different
configurations, and to do atomic upgrades and rollbacks.  To
understand how they work, it&#x2019;s useful to know a bit about how Nix
works.  In Nix, packages are stored in unique locations in the
<emphasis>Nix store</emphasis> (typically,
<filename>/nix/store</filename>).  For instance, a particular version
of the Subversion package might be stored in a directory
<filename>/nix/store/dpmvp969yhdqs7lm2r1a3gng7pyq6vy4-subversion-1.1.3/</filename>,
while another version might be stored in
<filename>/nix/store/5mq2jcn36ldlmh93yj1n8s9c95pj7c5s-subversion-1.1.2</filename>.
The long strings prefixed to the directory names are cryptographic
hashes<footnote><para>160-bit truncations of SHA-256 hashes encoded in
a base-32 notation, to be precise.</para></footnote> of
<emphasis>all</emphasis> inputs involved in building the package &#x2014;
sources, dependencies, compiler flags, and so on.  So if two
packages differ in any way, they end up in different locations in
the file system, so they don&#x2019;t interfere with each other.  <xref linkend="fig-user-environments"/> shows a part of a typical Nix
store.</para>

<figure xml:id="fig-user-environments"><title>User environments</title>
  <mediaobject>
    <imageobject>
      <imagedata fileref="../figures/user-environments.png" format="PNG"/>
    </imageobject>
  </mediaobject>
</figure>

<para>Of course, you wouldn&#x2019;t want to type

<screen>
$ /nix/store/dpmvp969yhdq...-subversion-1.1.3/bin/svn</screen>

every time you want to run Subversion.  Of course we could set up the
<envar>PATH</envar> environment variable to include the
<filename>bin</filename> directory of every package we want to use,
but this is not very convenient since changing <envar>PATH</envar>
doesn&#x2019;t take effect for already existing processes.  The solution Nix
uses is to create directory trees of symlinks to
<emphasis>activated</emphasis> packages.  These are called
<emphasis>user environments</emphasis> and they are packages
themselves (though automatically generated by
<command>nix-env</command>), so they too reside in the Nix store.  For
instance, in <xref linkend="fig-user-environments"/> the user
environment <filename>/nix/store/0c1p5z4kda11...-user-env</filename>
contains a symlink to just Subversion 1.1.2 (arrows in the figure
indicate symlinks).  This would be what we would obtain if we had done

<screen>
$ nix-env -i subversion</screen>

on a set of Nix expressions that contained Subversion 1.1.2.</para>

<para>This doesn&#x2019;t in itself solve the problem, of course; you
wouldn&#x2019;t want to type
<filename>/nix/store/0c1p5z4kda11...-user-env/bin/svn</filename>
either.  That&#x2019;s why there are symlinks outside of the store that point
to the user environments in the store; for instance, the symlinks
<filename>default-42-link</filename> and
<filename>default-43-link</filename> in the example.  These are called
<emphasis>generations</emphasis> since every time you perform a
<command>nix-env</command> operation, a new user environment is
generated based on the current one.  For instance, generation 43 was
created from generation 42 when we did

<screen>
$ nix-env -i subversion firefox</screen>

on a set of Nix expressions that contained Firefox and a new version
of Subversion.</para>

<para>Generations are grouped together into
<emphasis>profiles</emphasis> so that different users don&#x2019;t interfere
with each other if they don&#x2019;t want to.  For example:

<screen>
$ ls -l /nix/var/nix/profiles/
...
lrwxrwxrwx  1 eelco ... default-42-link -&gt; /nix/store/0c1p5z4kda11...-user-env
lrwxrwxrwx  1 eelco ... default-43-link -&gt; /nix/store/3aw2pdyx2jfc...-user-env
lrwxrwxrwx  1 eelco ... default -&gt; default-43-link</screen>

This shows a profile called <filename>default</filename>.  The file
<filename>default</filename> itself is actually a symlink that points
to the current generation.  When we do a <command>nix-env</command>
operation, a new user environment and generation link are created
based on the current one, and finally the <filename>default</filename>
symlink is made to point at the new generation.  This last step is
atomic on Unix, which explains how we can do atomic upgrades.  (Note
that the building/installing of new packages doesn&#x2019;t interfere in
any way with old packages, since they are stored in different
locations in the Nix store.)</para>

<para>If you find that you want to undo a <command>nix-env</command>
operation, you can just do

<screen>
$ nix-env --rollback</screen>

which will just make the current generation link point at the previous
link.  E.g., <filename>default</filename> would be made to point at
<filename>default-42-link</filename>.  You can also switch to a
specific generation:

<screen>
$ nix-env --switch-generation 43</screen>

which in this example would roll forward to generation 43 again.  You
can also see all available generations:

<screen>
$ nix-env --list-generations</screen></para>

<para>You generally wouldn&#x2019;t have
<filename>/nix/var/nix/profiles/<replaceable>some-profile</replaceable>/bin</filename>
in your <envar>PATH</envar>.  Rather, there is a symlink
<filename>~/.nix-profile</filename> that points to your current
profile.  This means that you should put
<filename>~/.nix-profile/bin</filename> in your <envar>PATH</envar>
(and indeed, that&#x2019;s what the initialisation script
<filename>/nix/etc/profile.d/nix.sh</filename> does).  This makes it
easier to switch to a different profile.  You can do that using the
command <command>nix-env --switch-profile</command>:

<screen>
$ nix-env --switch-profile /nix/var/nix/profiles/my-profile

$ nix-env --switch-profile /nix/var/nix/profiles/default</screen>

These commands switch to the <filename>my-profile</filename> and
default profile, respectively.  If the profile doesn&#x2019;t exist, it will
be created automatically.  You should be careful about storing a
profile in another location than the <filename>profiles</filename>
directory, since otherwise it might not be used as a root of the
garbage collector (see <xref linkend="sec-garbage-collection"/>).</para>

<para>All <command>nix-env</command> operations work on the profile
pointed to by <command>~/.nix-profile</command>, but you can override
this using the <option>--profile</option> option (abbreviation
<option>-p</option>):

<screen>
$ nix-env -p /nix/var/nix/profiles/other-profile -i subversion</screen>

This will <emphasis>not</emphasis> change the
<command>~/.nix-profile</command> symlink.</para>

</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-garbage-collection">

<title>Garbage Collection</title>

<para><command>nix-env</command> operations such as upgrades
(<option>-u</option>) and uninstall (<option>-e</option>) never
actually delete packages from the system.  All they do (as shown
above) is to create a new user environment that no longer contains
symlinks to the &#x201C;deleted&#x201D; packages.</para>

<para>Of course, since disk space is not infinite, unused packages
should be removed at some point.  You can do this by running the Nix
garbage collector.  It will remove from the Nix store any package
not used (directly or indirectly) by any generation of any
profile.</para>

<para>Note however that as long as old generations reference a
package, it will not be deleted.  After all, we wouldn&#x2019;t be able to
do a rollback otherwise.  So in order for garbage collection to be
effective, you should also delete (some) old generations.  Of course,
this should only be done if you are certain that you will not need to
roll back.</para>

<para>To delete all old (non-current) generations of your current
profile:

<screen>
$ nix-env --delete-generations old</screen>

Instead of <literal>old</literal> you can also specify a list of
generations, e.g.,

<screen>
$ nix-env --delete-generations 10 11 14</screen>

To delete all generations older than a specified number of days
(except the current generation), use the <literal>d</literal>
suffix. For example,

<screen>
$ nix-env --delete-generations 14d</screen>

deletes all generations older than two weeks.</para>

<para>After removing appropriate old generations you can run the
garbage collector as follows:

<screen>
$ nix-store --gc</screen>

The behaviour of the gargage collector is affected by the 
<literal>keep-derivations</literal> (default: true) and <literal>keep-outputs</literal>
(default: false) options in the Nix configuration file. The defaults will ensure
that all derivations that are build-time dependencies of garbage collector roots
will be kept and that all output paths that are runtime dependencies
will be kept as well. All other derivations or paths will be collected. 
(This is usually what you want, but while you are developing
it may make sense to keep outputs to ensure that rebuild times are quick.)

If you are feeling uncertain, you can also first view what files would
be deleted:

<screen>
$ nix-store --gc --print-dead</screen>

Likewise, the option <option>--print-live</option> will show the paths
that <emphasis>won&#x2019;t</emphasis> be deleted.</para>

<para>There is also a convenient little utility
<command>nix-collect-garbage</command>, which when invoked with the
<option>-d</option> (<option>--delete-old</option>) switch deletes all
old generations of all profiles in
<filename>/nix/var/nix/profiles</filename>.  So

<screen>
$ nix-collect-garbage -d</screen>

is a quick and easy way to clean up your system.</para>

<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-gc-roots">

<title>Garbage Collector Roots</title>

<para>The roots of the garbage collector are all store paths to which
there are symlinks in the directory
<filename><replaceable>prefix</replaceable>/nix/var/nix/gcroots</filename>.
For instance, the following command makes the path
<filename>/nix/store/d718ef...-foo</filename> a root of the collector:

<screen>
$ ln -s /nix/store/d718ef...-foo /nix/var/nix/gcroots/bar</screen>
	
That is, after this command, the garbage collector will not remove
<filename>/nix/store/d718ef...-foo</filename> or any of its
dependencies.</para>

<para>Subdirectories of
<filename><replaceable>prefix</replaceable>/nix/var/nix/gcroots</filename>
are also searched for symlinks.  Symlinks to non-store paths are
followed and searched for roots, but symlinks to non-store paths
<emphasis>inside</emphasis> the paths reached in that way are not
followed to prevent infinite recursion.</para>

</section>

</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-channels">

<title>Channels</title>

<para>If you want to stay up to date with a set of packages, it&#x2019;s not
very convenient to manually download the latest set of Nix expressions
for those packages and upgrade using <command>nix-env</command>.
Fortunately, there&#x2019;s a better way: <emphasis>Nix
channels</emphasis>.</para>

<para>A Nix channel is just a URL that points to a place that contains
a set of Nix expressions and a manifest.  Using the command <link linkend="sec-nix-channel"><command>nix-channel</command></link> you
can automatically stay up to date with whatever is available at that
URL.</para>
      
<para>To see the list of official NixOS channels, visit <link xlink:href="https://nixos.org/channels"/>.</para>

<para>You can &#x201C;subscribe&#x201D; to a channel using
<command>nix-channel --add</command>, e.g.,

<screen>
$ nix-channel --add https://nixos.org/channels/nixpkgs-unstable</screen>

subscribes you to a channel that always contains that latest version
of the Nix Packages collection.  (Subscribing really just means that
the URL is added to the file <filename>~/.nix-channels</filename>,
where it is read by subsequent calls to <command>nix-channel
--update</command>.) You can &#x201C;unsubscribe&#x201D; using <command>nix-channel
--remove</command>:

<screen>
$ nix-channel --remove nixpkgs
</screen>
</para>

<para>To obtain the latest Nix expressions available in a channel, do

<screen>
$ nix-channel --update</screen>

This downloads and unpacks the Nix expressions in every channel
(downloaded from <literal><replaceable>url</replaceable>/nixexprs.tar.bz2</literal>).
It also makes the union of each channel&#x2019;s Nix expressions available by
default to <command>nix-env</command> operations (via the symlink
<filename>~/.nix-defexpr/channels</filename>).  Consequently, you can
then say

<screen>
$ nix-env -u</screen>

to upgrade all packages in your profile to the latest versions
available in the subscribed channels.</para>

</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-sharing-packages">

<title>Sharing Packages Between Machines</title>

<para>Sometimes you want to copy a package from one machine to
another.  Or, you want to install some packages and you know that
another machine already has some or all of those packages or their
dependencies.  In that case there are mechanisms to quickly copy
packages between machines.</para>

<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-binary-cache-substituter">

<title>Serving a Nix store via HTTP</title>

<para>You can easily share the Nix store of a machine via HTTP. This
allows other machines to fetch store paths from that machine to speed
up installations. It uses the same <emphasis>binary cache</emphasis>
mechanism that Nix usually uses to fetch pre-built binaries from
<uri>https://cache.nixos.org</uri>.</para>

<para>The daemon that handles binary cache requests via HTTP,
<command>nix-serve</command>, is not part of the Nix distribution, but
you can install it from Nixpkgs:

<screen>
$ nix-env -i nix-serve
</screen>

You can then start the server, listening for HTTP connections on
whatever port you like:

<screen>
$ nix-serve -p 8080
</screen>

To check whether it works, try the following on the client:

<screen>
$ curl http://avalon:8080/nix-cache-info
</screen>

which should print something like:

<screen>
StoreDir: /nix/store
WantMassQuery: 1
Priority: 30
</screen>

</para>

<para>On the client side, you can tell Nix to use your binary cache
using <option>--option extra-binary-caches</option>, e.g.:

<screen>
$ nix-env -i firefox --option extra-binary-caches http://avalon:8080/
</screen>

The option <option>extra-binary-caches</option> tells Nix to use this
binary cache in addition to your default caches, such as
<uri>https://cache.nixos.org</uri>. Thus, for any path in the closure
of Firefox, Nix will first check if the path is available on the
server <literal>avalon</literal> or another binary caches. If not, it
will fall back to building from source.</para>

<para>You can also tell Nix to always use your binary cache by adding
a line to the <filename linkend="sec-conf-file">nix.conf</filename>
configuration file like this:

<programlisting>
binary-caches = http://avalon:8080/ https://cache.nixos.org/
</programlisting>

</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-copy-closure">

<title>Copying Closures Via SSH</title>

<para>The command <command linkend="sec-nix-copy-closure">nix-copy-closure</command> copies a Nix
store path along with all its dependencies to or from another machine
via the SSH protocol.  It doesn&#x2019;t copy store paths that are already
present on the target machine.  For example, the following command
copies Firefox with all its dependencies:

<screen>
$ nix-copy-closure --to alice@itchy.example.org $(type -p firefox)</screen>

See <xref linkend="sec-nix-copy-closure"/> for details.</para>

<para>With <command linkend="refsec-nix-store-export">nix-store
--export</command> and <command linkend="refsec-nix-store-import">nix-store --import</command> you can
write the closure of a store path (that is, the path and all its
dependencies) to a file, and then unpack that file into another Nix
store.  For example,

<screen>
$ nix-store --export $(nix-store -qR $(type -p firefox)) &gt; firefox.closure</screen>

writes the closure of Firefox to a file.  You can then copy this file
to another machine and install the closure:

<screen>
$ nix-store --import &lt; firefox.closure</screen>

Any store paths in the closure that are already present in the target
store are ignored.  It is also possible to pipe the export into
another command, e.g. to copy and install a closure directly to/on
another machine:

<screen>
$ nix-store --export $(nix-store -qR $(type -p firefox)) | bzip2 | \
    ssh alice@itchy.example.org "bunzip2 | nix-store --import"</screen>

However, <command>nix-copy-closure</command> is generally more
efficient because it only copies paths that are not already present in
the target Nix store.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-ssh-substituter">

<title>Serving a Nix store via SSH</title>

<para>You can tell Nix to automatically fetch needed binaries from a
remote Nix store via SSH. For example, the following installs Firefox,
automatically fetching any store paths in Firefox&#x2019;s closure if they
are available on the server <literal>avalon</literal>:

<screen>
$ nix-env -i firefox --substituters ssh://alice@avalon
</screen>

This works similar to the binary cache substituter that Nix usually
uses, only using SSH instead of HTTP: if a store path
<literal>P</literal> is needed, Nix will first check if it&#x2019;s available
in the Nix store on <literal>avalon</literal>. If not, it will fall
back to using the binary cache substituter, and then to building from
source.</para>

<note><para>The SSH substituter currently does not allow you to enter
an SSH passphrase interactively. Therefore, you should use
<command>ssh-add</command> to load the decrypted private key into
<command>ssh-agent</command>.</para></note>

<para>You can also copy the closure of some store path, without
installing it into your profile, e.g.

<screen>
$ nix-store -r /nix/store/m85bxg&#x2026;-firefox-34.0.5 --substituters ssh://alice@avalon
</screen>

This is essentially equivalent to doing

<screen>
$ nix-copy-closure --from alice@avalon /nix/store/m85bxg&#x2026;-firefox-34.0.5
</screen>

</para>

<para>You can use SSH&#x2019;s <emphasis>forced command</emphasis> feature to
set up a restricted user account for SSH substituter access, allowing
read-only access to the local Nix store, but nothing more. For
example, add the following lines to <filename>sshd_config</filename>
to restrict the user <literal>nix-ssh</literal>:

<programlisting>
Match User nix-ssh
  AllowAgentForwarding no
  AllowTcpForwarding no
  PermitTTY no
  PermitTunnel no
  X11Forwarding no
  ForceCommand nix-store --serve
Match All
</programlisting>

On NixOS, you can accomplish the same by adding the following to your
<filename>configuration.nix</filename>:

<programlisting>
nix.sshServe.enable = true;
nix.sshServe.keys = [ "ssh-dss AAAAB3NzaC1k... bob@example.org" ];
</programlisting>

where the latter line lists the public keys of users that are allowed
to connect.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-s3-substituter">

<title>Serving a Nix store via AWS S3 or S3-compatible Service</title>

<para>Nix has built-in support for storing and fetching store paths
from Amazon S3 and S3 compatible services. This uses the same
<emphasis>binary</emphasis> cache mechanism that Nix usually uses to
fetch prebuilt binaries from <uri>cache.nixos.org</uri>.</para>

<para>The following options can be specified as URL parameters to
the S3 URL:</para>

<variablelist>
  <varlistentry><term><literal>profile</literal></term>
  <listitem>
    <para>
      The name of the AWS configuration profile to use. By default
      Nix will use the <literal>default</literal> profile.
    </para>
  </listitem>
  </varlistentry>

  <varlistentry><term><literal>region</literal></term>
  <listitem>
    <para>
      The region of the S3 bucket. <literal>us&#x2013;east-1</literal> by
      default.
    </para>

    <para>
      If your bucket is not in <literal>us&#x2013;east-1</literal>, you
      should always explicitly specify the region parameter.
    </para>
  </listitem>
  </varlistentry>

  <varlistentry><term><literal>endpoint</literal></term>
  <listitem>
    <para>
      The URL to your S3-compatible service, for when not using
      Amazon S3. Do not specify this value if you're using Amazon
      S3.
    </para>
    <note><para>This endpoint must support HTTPS and will use
    path-based addressing instead of virtual host based
    addressing.</para></note>
  </listitem>
  </varlistentry>

  <varlistentry><term><literal>scheme</literal></term>
  <listitem>
    <para>
      The scheme used for S3 requests, <literal>https</literal>
      (default) or <literal>http</literal>.  This option allows you to
      disable HTTPS for binary caches which don't support it.
    </para>
    <note><para>HTTPS should be used if the cache might contain
    sensitive information.</para></note>
  </listitem>
  </varlistentry>
</variablelist>

<para>In this example we will use the bucket named
<literal>example-nix-cache</literal>.</para>

<section xml:id="ssec-s3-substituter-anonymous-reads">
  <title>Anonymous Reads to your S3-compatible binary cache</title>

  <para>If your binary cache is publicly accessible and does not
  require authentication, the simplest and easiest way to use Nix with
  your S3 compatible binary cache is to use the HTTP URL for that
  cache.</para>

  <para>For AWS S3 the binary cache URL for example bucket will be
  exactly <uri>https://example-nix-cache.s3.amazonaws.com</uri> or
  <uri>s3://example-nix-cache</uri>. For S3 compatible binary caches,
  consult that cache's documentation.</para>

  <para>Your bucket will need the following bucket policy:</para>

  <programlisting><![CDATA[
{
    "Id": "DirectReads",
    "Version": "2012-10-17",
    "Statement": [
        {
            "Sid": "AllowDirectReads",
            "Action": [
                "s3:GetObject",
                "s3:GetBucketLocation"
            ],
            "Effect": "Allow",
            "Resource": [
                "arn:aws:s3:::example-nix-cache",
                "arn:aws:s3:::example-nix-cache/*"
            ],
            "Principal": "*"
        }
    ]
}
]]></programlisting>
</section>

<section xml:id="ssec-s3-substituter-authenticated-reads">
  <title>Authenticated Reads to your S3 binary cache</title>

  <para>For AWS S3 the binary cache URL for example bucket will be
  exactly <uri>s3://example-nix-cache</uri>.</para>

  <para>Nix will use the <link xlink:href="https://docs.aws.amazon.com/sdk-for-cpp/v1/developer-guide/credentials.html">default
  credential provider chain</link> for authenticating requests to
  Amazon S3.</para>

  <para>Nix supports authenticated reads from Amazon S3 and S3
  compatible binary caches.</para>

  <para>Your bucket will need a bucket policy allowing the desired
  users to perform the <literal>s3:GetObject</literal> and
  <literal>s3:GetBucketLocation</literal> action on all objects in the
  bucket. The anonymous policy in <xref linkend="ssec-s3-substituter-anonymous-reads"/> can be updated to
  have a restricted <literal>Principal</literal> to support
  this.</para>
</section>


<section xml:id="ssec-s3-substituter-authenticated-writes">
  <title>Authenticated Writes to your S3-compatible binary cache</title>

  <para>Nix support fully supports writing to Amazon S3 and S3
  compatible buckets. The binary cache URL for our example bucket will
  be <uri>s3://example-nix-cache</uri>.</para>

  <para>Nix will use the <link xlink:href="https://docs.aws.amazon.com/sdk-for-cpp/v1/developer-guide/credentials.html">default
  credential provider chain</link> for authenticating requests to
  Amazon S3.</para>

  <para>Your account will need the following IAM policy to
  upload to the cache:</para>

  <programlisting><![CDATA[
{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Sid": "UploadToCache",
      "Effect": "Allow",
      "Action": [
        "s3:AbortMultipartUpload",
        "s3:GetBucketLocation",
        "s3:GetObject",
        "s3:ListBucket",
        "s3:ListBucketMultipartUploads",
        "s3:ListMultipartUploadParts",
        "s3:PutObject"
      ],
      "Resource": [
        "arn:aws:s3:::example-nix-cache",
        "arn:aws:s3:::example-nix-cache/*"
      ]
    }
  ]
}
]]></programlisting>


  <example><title>Uploading with a specific credential profile for Amazon S3</title>
    <para><command>nix copy --to 's3://example-nix-cache?profile=cache-upload&amp;region=eu-west-2' nixpkgs.hello</command></para>
  </example>

  <example><title>Uploading to an S3-Compatible Binary Cache</title>
    <para><command>nix copy --to 's3://example-nix-cache?profile=cache-upload&amp;scheme=https&amp;endpoint=minio.example.com' nixpkgs.hello</command></para>
  </example>
</section>
</section>

</chapter>

</part>
  <part xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="chap-writing-nix-expressions" xml:base="expressions/writing-nix-expressions.xml">

<title>Writing Nix Expressions</title>

<partintro>
<para>This chapter shows you how to write Nix expressions, which 
instruct Nix how to build packages.  It starts with a
simple example (a Nix expression for GNU Hello), and then moves
on to a more in-depth look at the Nix expression language.</para>

<note><para>This chapter is mostly about the Nix expression language.
For more extensive information on adding packages to the Nix Packages
collection (such as functions in the standard environment and coding
conventions), please consult <link xlink:href="http://nixos.org/nixpkgs/manual/">its
manual</link>.</para></note>
</partintro>

<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-simple-expression">

<title>A Simple Nix Expression</title>

<para>This section shows how to add and test the <link xlink:href="http://www.gnu.org/software/hello/hello.html">GNU Hello
package</link> to the Nix Packages collection.  Hello is a program
that prints out the text <quote>Hello, world!</quote>.</para>

<para>To add a package to the Nix Packages collection, you generally
need to do three things:

<orderedlist>

  <listitem><para>Write a Nix expression for the package.  This is a
  file that describes all the inputs involved in building the package,
  such as dependencies, sources, and so on.</para></listitem>

  <listitem><para>Write a <emphasis>builder</emphasis>.  This is a
  shell script<footnote><para>In fact, it can be written in any
  language, but typically it's a <command>bash</command> shell
  script.</para></footnote> that actually builds the package from
  the inputs.</para></listitem>

  <listitem><para>Add the package to the file
  <filename>pkgs/top-level/all-packages.nix</filename>.  The Nix
  expression written in the first step is a
  <emphasis>function</emphasis>; it requires other packages in order
  to build it.  In this step you put it all together, i.e., you call
  the function with the right arguments to build the actual
  package.</para></listitem>

</orderedlist>

</para>

<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-expression-syntax">

<title>Expression Syntax</title>

<example xml:id="ex-hello-nix"><title>Nix expression for GNU Hello
(<filename>default.nix</filename>)</title>
<programlisting>
{ stdenv, fetchurl, perl }: <co xml:id="ex-hello-nix-co-1"/>

stdenv.mkDerivation { <co xml:id="ex-hello-nix-co-2"/>
  name = "hello-2.1.1"; <co xml:id="ex-hello-nix-co-3"/>
  builder = ./builder.sh; <co xml:id="ex-hello-nix-co-4"/>
  src = fetchurl { <co xml:id="ex-hello-nix-co-5"/>
    url = "ftp://ftp.nluug.nl/pub/gnu/hello/hello-2.1.1.tar.gz";
    sha256 = "1md7jsfd8pa45z73bz1kszpp01yw6x5ljkjk2hx7wl800any6465";
  };
  inherit perl; <co xml:id="ex-hello-nix-co-6"/>
}</programlisting>
</example>

<para><xref linkend="ex-hello-nix"/> shows a Nix expression for GNU
Hello.  It's actually already in the Nix Packages collection in
<filename>pkgs/applications/misc/hello/ex-1/default.nix</filename>.
It is customary to place each package in a separate directory and call
the single Nix expression in that directory
<filename>default.nix</filename>.  The file has the following elements
(referenced from the figure by number):

<calloutlist>

  <callout arearefs="ex-hello-nix-co-1">

    <para>This states that the expression is a
    <emphasis>function</emphasis> that expects to be called with three
    arguments: <varname>stdenv</varname>, <varname>fetchurl</varname>,
    and <varname>perl</varname>.  They are needed to build Hello, but
    we don't know how to build them here; that's why they are function
    arguments.  <varname>stdenv</varname> is a package that is used
    by almost all Nix Packages packages; it provides a
    <quote>standard</quote> environment consisting of the things you
    would expect in a basic Unix environment: a C/C++ compiler (GCC,
    to be precise), the Bash shell, fundamental Unix tools such as
    <command>cp</command>, <command>grep</command>,
    <command>tar</command>, etc.  <varname>fetchurl</varname> is a
    function that downloads files.  <varname>perl</varname> is the
    Perl interpreter.</para>

    <para>Nix functions generally have the form <literal>{ x, y, ...,
    z }: e</literal> where <varname>x</varname>, <varname>y</varname>,
    etc. are the names of the expected arguments, and where
    <replaceable>e</replaceable> is the body of the function.  So
    here, the entire remainder of the file is the body of the
    function; when given the required arguments, the body should
    describe how to build an instance of the Hello package.</para>

  </callout>

  <callout arearefs="ex-hello-nix-co-2">

    <para>So we have to build a package.  Building something from
    other stuff is called a <emphasis>derivation</emphasis> in Nix (as
    opposed to sources, which are built by humans instead of
    computers).  We perform a derivation by calling
    <varname>stdenv.mkDerivation</varname>.
    <varname>mkDerivation</varname> is a function provided by
    <varname>stdenv</varname> that builds a package from a set of
    <emphasis>attributes</emphasis>.  A set is just a list of
    key/value pairs where each key is a string and each value is an
    arbitrary Nix expression.  They take the general form <literal>{
    <replaceable>name1</replaceable> =
    <replaceable>expr1</replaceable>; <replaceable>...</replaceable>
    <replaceable>nameN</replaceable> =
    <replaceable>exprN</replaceable>; }</literal>.</para>

  </callout>

  <callout arearefs="ex-hello-nix-co-3">

    <para>The attribute <varname>name</varname> specifies the symbolic
    name and version of the package.  Nix doesn't really care about
    these things, but they are used by for instance <command>nix-env
    -q</command> to show a <quote>human-readable</quote> name for
    packages.  This attribute is required by
    <varname>mkDerivation</varname>.</para>

  </callout>

  <callout arearefs="ex-hello-nix-co-4">

    <para>The attribute <varname>builder</varname> specifies the
    builder.  This attribute can sometimes be omitted, in which case
    <varname>mkDerivation</varname> will fill in a default builder
    (which does a <literal>configure; make; make install</literal>, in
    essence).  Hello is sufficiently simple that the default builder
    would suffice, but in this case, we will show an actual builder
    for educational purposes.  The value
    <command>./builder.sh</command> refers to the shell script shown
    in <xref linkend="ex-hello-builder"/>, discussed below.</para>

  </callout>

  <callout arearefs="ex-hello-nix-co-5">

    <para>The builder has to know what the sources of the package
    are.  Here, the attribute <varname>src</varname> is bound to the
    result of a call to the <command>fetchurl</command> function.
    Given a URL and a SHA-256 hash of the expected contents of the file
    at that URL, this function builds a derivation that downloads the
    file and checks its hash.  So the sources are a dependency that
    like all other dependencies is built before Hello itself is
    built.</para>

    <para>Instead of <varname>src</varname> any other name could have
    been used, and in fact there can be any number of sources (bound
    to different attributes).  However, <varname>src</varname> is
    customary, and it's also expected by the default builder (which we
    don't use in this example).</para>

  </callout>

  <callout arearefs="ex-hello-nix-co-6">

    <para>Since the derivation requires Perl, we have to pass the
    value of the <varname>perl</varname> function argument to the
    builder.  All attributes in the set are actually passed as
    environment variables to the builder, so declaring an attribute

    <programlisting>
perl = perl;</programlisting>

    will do the trick: it binds an attribute <varname>perl</varname>
    to the function argument which also happens to be called
    <varname>perl</varname>.  However, it looks a bit silly, so there
    is a shorter syntax.  The <literal>inherit</literal> keyword
    causes the specified attributes to be bound to whatever variables
    with the same name happen to be in scope.</para>

  </callout>

</calloutlist>

</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-build-script">

<title>Build Script</title>

<example xml:id="ex-hello-builder"><title>Build script for GNU Hello
(<filename>builder.sh</filename>)</title>
<programlisting>
source $stdenv/setup <co xml:id="ex-hello-builder-co-1"/>

PATH=$perl/bin:$PATH <co xml:id="ex-hello-builder-co-2"/>

tar xvfz $src <co xml:id="ex-hello-builder-co-3"/>
cd hello-*
./configure --prefix=$out <co xml:id="ex-hello-builder-co-4"/>
make <co xml:id="ex-hello-builder-co-5"/>
make install</programlisting>
</example>

<para><xref linkend="ex-hello-builder"/> shows the builder referenced
from Hello's Nix expression (stored in
<filename>pkgs/applications/misc/hello/ex-1/builder.sh</filename>).
The builder can actually be made a lot shorter by using the
<emphasis>generic builder</emphasis> functions provided by
<varname>stdenv</varname>, but here we write out the build steps to
elucidate what a builder does.  It performs the following
steps:</para>

<calloutlist>

  <callout arearefs="ex-hello-builder-co-1">

    <para>When Nix runs a builder, it initially completely clears the
    environment (except for the attributes declared in the
    derivation).  For instance, the <envar>PATH</envar> variable is
    empty<footnote><para>Actually, it's initialised to
    <filename>/path-not-set</filename> to prevent Bash from setting it
    to a default value.</para></footnote>.  This is done to prevent
    undeclared inputs from being used in the build process.  If for
    example the <envar>PATH</envar> contained
    <filename>/usr/bin</filename>, then you might accidentally use
    <filename>/usr/bin/gcc</filename>.</para>

    <para>So the first step is to set up the environment.  This is
    done by calling the <filename>setup</filename> script of the
    standard environment.  The environment variable
    <envar>stdenv</envar> points to the location of the standard
    environment being used.  (It wasn't specified explicitly as an
    attribute in <xref linkend="ex-hello-nix"/>, but
    <varname>mkDerivation</varname> adds it automatically.)</para>

  </callout>

  <callout arearefs="ex-hello-builder-co-2">

    <para>Since Hello needs Perl, we have to make sure that Perl is in
    the <envar>PATH</envar>.  The <envar>perl</envar> environment
    variable points to the location of the Perl package (since it
    was passed in as an attribute to the derivation), so
    <filename><replaceable>$perl</replaceable>/bin</filename> is the
    directory containing the Perl interpreter.</para>

  </callout>

  <callout arearefs="ex-hello-builder-co-3">

    <para>Now we have to unpack the sources.  The
    <varname>src</varname> attribute was bound to the result of
    fetching the Hello source tarball from the network, so the
    <envar>src</envar> environment variable points to the location in
    the Nix store to which the tarball was downloaded.  After
    unpacking, we <command>cd</command> to the resulting source
    directory.</para>

    <para>The whole build is performed in a temporary directory
    created in <varname>/tmp</varname>, by the way.  This directory is
    removed after the builder finishes, so there is no need to clean
    up the sources afterwards.  Also, the temporary directory is
    always newly created, so you don't have to worry about files from
    previous builds interfering with the current build.</para>

  </callout>

  <callout arearefs="ex-hello-builder-co-4">

    <para>GNU Hello is a typical Autoconf-based package, so we first
    have to run its <filename>configure</filename> script.  In Nix
    every package is stored in a separate location in the Nix store,
    for instance
    <filename>/nix/store/9a54ba97fb71b65fda531012d0443ce2-hello-2.1.1</filename>.
    Nix computes this path by cryptographically hashing all attributes
    of the derivation.  The path is passed to the builder through the
    <envar>out</envar> environment variable.  So here we give
    <filename>configure</filename> the parameter
    <literal>--prefix=$out</literal> to cause Hello to be installed in
    the expected location.</para>

  </callout>

  <callout arearefs="ex-hello-builder-co-5">

    <para>Finally we build Hello (<literal>make</literal>) and install
    it into the location specified by <envar>out</envar>
    (<literal>make install</literal>).</para>

  </callout>

</calloutlist>

<para>If you are wondering about the absence of error checking on the
result of various commands called in the builder: this is because the
shell script is evaluated with Bash's <option>-e</option> option,
which causes the script to be aborted if any command fails without an
error check.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-arguments">

<title>Arguments and Variables</title>

<example xml:id="ex-hello-composition">

<title>Composing GNU Hello
(<filename>all-packages.nix</filename>)</title>
<programlisting>
...

rec { <co xml:id="ex-hello-composition-co-1"/>

  hello = import ../applications/misc/hello/ex-1 <co xml:id="ex-hello-composition-co-2"/> { <co xml:id="ex-hello-composition-co-3"/>
    inherit fetchurl stdenv perl;
  };

  perl = import ../development/interpreters/perl { <co xml:id="ex-hello-composition-co-4"/>
    inherit fetchurl stdenv;
  };

  fetchurl = import ../build-support/fetchurl {
    inherit stdenv; ...
  };

  stdenv = ...;

}
</programlisting>
</example>

<para>The Nix expression in <xref linkend="ex-hello-nix"/> is a
function; it is missing some arguments that have to be filled in
somewhere.  In the Nix Packages collection this is done in the file
<filename>pkgs/top-level/all-packages.nix</filename>, where all
Nix expressions for packages are imported and called with the
appropriate arguments.  <xref linkend="ex-hello-composition"/> shows
some fragments of
<filename>all-packages.nix</filename>.</para>

<calloutlist>

  <callout arearefs="ex-hello-composition-co-1">

    <para>This file defines a set of attributes, all of which are
    concrete derivations (i.e., not functions).  In fact, we define a
    <emphasis>mutually recursive</emphasis> set of attributes.  That
    is, the attributes can refer to each other.  This is precisely
    what we want since we want to <quote>plug</quote> the
    various packages into each other.</para>

  </callout>

  <callout arearefs="ex-hello-composition-co-2">

    <para>Here we <emphasis>import</emphasis> the Nix expression for
    GNU Hello.  The import operation just loads and returns the
    specified Nix expression. In fact, we could just have put the
    contents of <xref linkend="ex-hello-nix"/> in
    <filename>all-packages.nix</filename> at this point.  That
    would be completely equivalent, but it would make the file rather
    bulky.</para>

    <para>Note that we refer to
    <filename>../applications/misc/hello/ex-1</filename>, not
    <filename>../applications/misc/hello/ex-1/default.nix</filename>.
    When you try to import a directory, Nix automatically appends
    <filename>/default.nix</filename> to the file name.</para>

  </callout>

  <callout arearefs="ex-hello-composition-co-3">

    <para>This is where the actual composition takes place.  Here we
    <emphasis>call</emphasis> the function imported from
    <filename>../applications/misc/hello/ex-1</filename> with a set
    containing the things that the function expects, namely
    <varname>fetchurl</varname>, <varname>stdenv</varname>, and
    <varname>perl</varname>.  We use inherit again to use the
    attributes defined in the surrounding scope (we could also have
    written <literal>fetchurl = fetchurl;</literal>, etc.).</para>

    <para>The result of this function call is an actual derivation
    that can be built by Nix (since when we fill in the arguments of
    the function, what we get is its body, which is the call to
    <varname>stdenv.mkDerivation</varname> in <xref linkend="ex-hello-nix"/>).</para>

    <note><para>Nixpkgs has a convenience function
    <function>callPackage</function> that imports and calls a
    function, filling in any missing arguments by passing the
    corresponding attribute from the Nixpkgs set, like this:

<programlisting>
hello = callPackage ../applications/misc/hello/ex-1 { };
</programlisting>

    If necessary, you can set or override arguments:

<programlisting>
hello = callPackage ../applications/misc/hello/ex-1 { stdenv = myStdenv; };
</programlisting>

    </para></note>

  </callout>

  <callout arearefs="ex-hello-composition-co-4">

    <para>Likewise, we have to instantiate Perl,
    <varname>fetchurl</varname>, and the standard environment.</para>

  </callout>

</calloutlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-building-simple">

<title>Building and Testing</title>

<para>You can now try to build Hello.  Of course, you could do
<literal>nix-env -i hello</literal>, but you may not want to install a
possibly broken package just yet.  The best way to test the package is by
using the command <command linkend="sec-nix-build">nix-build</command>,
which builds a Nix expression and creates a symlink named
<filename>result</filename> in the current directory:

<screen>
$ nix-build -A hello
building path `/nix/store/632d2b22514d...-hello-2.1.1'
hello-2.1.1/
hello-2.1.1/intl/
hello-2.1.1/intl/ChangeLog
<replaceable>...</replaceable>

$ ls -l result
lrwxrwxrwx ... 2006-09-29 10:43 result -&gt; /nix/store/632d2b22514d...-hello-2.1.1

$ ./result/bin/hello
Hello, world!</screen>

The <link linkend="opt-attr"><option>-A</option></link> option selects
the <literal>hello</literal> attribute.  This is faster than using the
symbolic package name specified by the <literal>name</literal>
attribute (which also happens to be <literal>hello</literal>) and is
unambiguous (there can be multiple packages with the symbolic name
<literal>hello</literal>, but there can be only one attribute in a set
named <literal>hello</literal>).</para>

<para><command>nix-build</command> registers the
<filename>./result</filename> symlink as a garbage collection root, so
unless and until you delete the <filename>./result</filename> symlink,
the output of the build will be safely kept on your system.  You can
use <command>nix-build</command>&#x2019;s <option linkend="opt-out-link">-o</option> switch to give the symlink another
name.</para>

<para>Nix has transactional semantics.  Once a build finishes
successfully, Nix makes a note of this in its database: it registers
that the path denoted by <envar>out</envar> is now
<quote>valid</quote>.  If you try to build the derivation again, Nix
will see that the path is already valid and finish immediately.  If a
build fails, either because it returns a non-zero exit code, because
Nix or the builder are killed, or because the machine crashes, then
the output paths will not be registered as valid.  If you try to build
the derivation again, Nix will remove the output paths if they exist
(e.g., because the builder died half-way through <literal>make
install</literal>) and try again.  Note that there is no
<quote>negative caching</quote>: Nix doesn't remember that a build
failed, and so a failed build can always be repeated.  This is because
Nix cannot distinguish between permanent failures (e.g., a compiler
error due to a syntax error in the source) and transient failures
(e.g., a disk full condition).</para>

<para>Nix also performs locking.  If you run multiple Nix builds
simultaneously, and they try to build the same derivation, the first
Nix instance that gets there will perform the build, while the others
block (or perform other derivations if available) until the build
finishes:

<screen>
$ nix-build -A hello
waiting for lock on `/nix/store/0h5b7hp8d4hqfrw8igvx97x1xawrjnac-hello-2.1.1x'</screen>

So it is always safe to run multiple instances of Nix in parallel
(which isn&#x2019;t the case with, say, <command>make</command>).</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-generic-builder">

<title>Generic Builder Syntax</title>

<para>Recall from <xref linkend="ex-hello-builder"/> that the builder
looked something like this:

<programlisting>
PATH=$perl/bin:$PATH
tar xvfz $src
cd hello-*
./configure --prefix=$out
make
make install</programlisting>

The builders for almost all Unix packages look like this &#x2014; set up some
environment variables, unpack the sources, configure, build, and
install.  For this reason the standard environment provides some Bash
functions that automate the build process.  A builder using the
generic build facilities in shown in <xref linkend="ex-hello-builder2"/>.</para>

<example xml:id="ex-hello-builder2"><title>Build script using the generic
build functions</title>
<programlisting>
buildInputs="$perl" <co xml:id="ex-hello-builder2-co-1"/>

source $stdenv/setup <co xml:id="ex-hello-builder2-co-2"/>

genericBuild <co xml:id="ex-hello-builder2-co-3"/></programlisting>
</example>

<calloutlist>

  <callout arearefs="ex-hello-builder2-co-1">

    <para>The <envar>buildInputs</envar> variable tells
    <filename>setup</filename> to use the indicated packages as
    <quote>inputs</quote>.  This means that if a package provides a
    <filename>bin</filename> subdirectory, it's added to
    <envar>PATH</envar>; if it has a <filename>include</filename>
    subdirectory, it's added to GCC's header search path; and so
    on.<footnote><para>How does it work? <filename>setup</filename>
    tries to source the file
    <filename><replaceable>pkg</replaceable>/nix-support/setup-hook</filename>
    of all dependencies.  These &#x201C;setup hooks&#x201D; can then set up whatever
    environment variables they want; for instance, the setup hook for
    Perl sets the <envar>PERL5LIB</envar> environment variable to
    contain the <filename>lib/site_perl</filename> directories of all
    inputs.</para></footnote>
    </para>

  </callout>

  <callout arearefs="ex-hello-builder2-co-2">

    <para>The function <function>genericBuild</function> is defined in
    the file <literal>$stdenv/setup</literal>.</para>

  </callout>

  <callout arearefs="ex-hello-builder2-co-3">

    <para>The final step calls the shell function
    <function>genericBuild</function>, which performs the steps that
    were done explicitly in <xref linkend="ex-hello-builder"/>.  The
    generic builder is smart enough to figure out whether to unpack
    the sources using <command>gzip</command>,
    <command>bzip2</command>, etc.  It can be customised in many ways;
    see the Nixpkgs manual for details.</para>

  </callout>

</calloutlist>

<para>Discerning readers will note that the
<envar>buildInputs</envar> could just as well have been set in the Nix
expression, like this:

<programlisting>
  buildInputs = [ perl ];</programlisting>

The <varname>perl</varname> attribute can then be removed, and the
builder becomes even shorter:

<programlisting>
source $stdenv/setup
genericBuild</programlisting>

In fact, <varname>mkDerivation</varname> provides a default builder
that looks exactly like that, so it is actually possible to omit the
builder for Hello entirely.</para>

</section>

</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-expression-language">

<title>Nix Expression Language</title>

<para>The Nix expression language is a pure, lazy, functional
language.  Purity means that operations in the language don't have
side-effects (for instance, there is no variable assignment).
Laziness means that arguments to functions are evaluated only when
they are needed.  Functional means that functions are
<quote>normal</quote> values that can be passed around and manipulated
in interesting ways.  The language is not a full-featured, general
purpose language.  Its main job is to describe packages,
compositions of packages, and the variability within
packages.</para>

<para>This section presents the various features of the
language.</para>

<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-values">

<title>Values</title>


<simplesect><title>Simple Values</title>

<para>Nix has the following basic data types:

<itemizedlist>

  <listitem>

    <para><emphasis>Strings</emphasis> can be written in three
    ways.</para>

    <para>The most common way is to enclose the string between double
    quotes, e.g., <literal>"foo bar"</literal>.  Strings can span
    multiple lines.  The special characters <literal>"</literal> and
    <literal>\</literal> and the character sequence
    <literal>${</literal> must be escaped by prefixing them with a
    backslash (<literal>\</literal>).  Newlines, carriage returns and
    tabs can be written as <literal>\n</literal>,
    <literal>\r</literal> and <literal>\t</literal>,
    respectively.</para>

    <para>You can include the result of an expression into a string by
    enclosing it in
    <literal>${<replaceable>...</replaceable>}</literal>, a feature
    known as <emphasis>antiquotation</emphasis>.  The enclosed
    expression must evaluate to something that can be coerced into a
    string (meaning that it must be a string, a path, or a
    derivation).  For instance, rather than writing

<programlisting>
"--with-freetype2-library=" + freetype + "/lib"</programlisting>

    (where <varname>freetype</varname> is a derivation), you can
    instead write the more natural

<programlisting>
"--with-freetype2-library=${freetype}/lib"</programlisting>

    The latter is automatically translated to the former.  A more
    complicated example (from the Nix expression for <link xlink:href="http://www.trolltech.com/products/qt">Qt</link>):

<programlisting>
configureFlags = "
  -system-zlib -system-libpng -system-libjpeg
  ${if openglSupport then "-dlopen-opengl
    -L${mesa}/lib -I${mesa}/include
    -L${libXmu}/lib -I${libXmu}/include" else ""}
  ${if threadSupport then "-thread" else "-no-thread"}
";</programlisting>

    Note that Nix expressions and strings can be arbitrarily nested;
    in this case the outer string contains various antiquotations that
    themselves contain strings (e.g., <literal>"-thread"</literal>),
    some of which in turn contain expressions (e.g.,
    <literal>${mesa}</literal>).</para>

    <para>The second way to write string literals is as an
    <emphasis>indented string</emphasis>, which is enclosed between
    pairs of <emphasis>double single-quotes</emphasis>, like so:

<programlisting>
''
  This is the first line.
  This is the second line.
    This is the third line.
''</programlisting>

    This kind of string literal intelligently strips indentation from
    the start of each line.  To be precise, it strips from each line a
    number of spaces equal to the minimal indentation of the string as
    a whole (disregarding the indentation of empty lines).  For
    instance, the first and second line are indented two space, while
    the third line is indented four spaces.  Thus, two spaces are
    stripped from each line, so the resulting string is

<programlisting>
"This is the first line.\nThis is the second line.\n  This is the third line.\n"</programlisting>

    </para>

    <para>Note that the whitespace and newline following the opening
    <literal>''</literal> is ignored if there is no non-whitespace
    text on the initial line.</para>

    <para>Antiquotation
    (<literal>${<replaceable>expr</replaceable>}</literal>) is
    supported in indented strings.</para>

    <para>Since <literal>${</literal> and <literal>''</literal> have
    special meaning in indented strings, you need a way to quote them.
    <literal>$</literal> can be escaped by prefixing it with
    <literal>''</literal> (that is, two single quotes), i.e.,
    <literal>''$</literal>. <literal>''</literal> can be escaped by
    prefixing it with <literal>'</literal>, i.e.,
    <literal>'''</literal>. <literal>$</literal> removes any special meaning
    from the following <literal>$</literal>. Linefeed, carriage-return and tab
    characters can be written as <literal>''\n</literal>,
    <literal>''\r</literal>, <literal>''\t</literal>, and <literal>''\</literal>
    escapes any other character.

    </para>

    <para>Indented strings are primarily useful in that they allow
    multi-line string literals to follow the indentation of the
    enclosing Nix expression, and that less escaping is typically
    necessary for strings representing languages such as shell scripts
    and configuration files because <literal>''</literal> is much less
    common than <literal>"</literal>.  Example:

<programlisting>
stdenv.mkDerivation {
  <replaceable>...</replaceable>
  postInstall =
    ''
      mkdir $out/bin $out/etc
      cp foo $out/bin
      echo "Hello World" &gt; $out/etc/foo.conf
      ${if enableBar then "cp bar $out/bin" else ""}
    '';
  <replaceable>...</replaceable>
}
</programlisting>

    </para>

    <para>Finally, as a convenience, <emphasis>URIs</emphasis> as
    defined in appendix B of <link xlink:href="http://www.ietf.org/rfc/rfc2396.txt">RFC 2396</link>
    can be written <emphasis>as is</emphasis>, without quotes.  For
    instance, the string
    <literal>"http://example.org/foo.tar.bz2"</literal>
    can also be written as
    <literal>http://example.org/foo.tar.bz2</literal>.</para>

  </listitem>

  <listitem><para>Numbers, which can be <emphasis>integers</emphasis> (like
  <literal>123</literal>) or <emphasis>floating point</emphasis> (like
  <literal>123.43</literal> or <literal>.27e13</literal>).</para>

  <para>Numbers are type-compatible: pure integer operations will always
  return integers, whereas any operation involving at least one floating point
  number will have a floating point number as a result.</para></listitem>

  <listitem><para><emphasis>Paths</emphasis>, e.g.,
  <filename>/bin/sh</filename> or <filename>./builder.sh</filename>.
  A path must contain at least one slash to be recognised as such; for
  instance, <filename>builder.sh</filename> is not a
  path<footnote><para>It's parsed as an expression that selects the
  attribute <varname>sh</varname> from the variable
  <varname>builder</varname>.</para></footnote>.  If the file name is
  relative, i.e., if it does not begin with a slash, it is made
  absolute at parse time relative to the directory of the Nix
  expression that contained it.  For instance, if a Nix expression in
  <filename>/foo/bar/bla.nix</filename> refers to
  <filename>../xyzzy/fnord.nix</filename>, the absolute path is
  <filename>/foo/xyzzy/fnord.nix</filename>.</para>

  <para>If the first component of a path is a <literal>~</literal>,
  it is interpreted as if the rest of the path were relative to the
  user's home directory. e.g. <filename>~/foo</filename> would be
  equivalent to <filename>/home/edolstra/foo</filename> for a user
  whose home directory is <filename>/home/edolstra</filename>.
  </para>

  <para>Paths can also be specified between angle brackets, e.g.
  <literal>&lt;nixpkgs&gt;</literal>. This means that the directories
  listed in the environment variable
  <envar linkend="env-NIX_PATH">NIX_PATH</envar> will be searched
  for the given file or directory name.
  </para>

  </listitem>

  <listitem><para><emphasis>Booleans</emphasis> with values
  <literal>true</literal> and
  <literal>false</literal>.</para></listitem>

  <listitem><para>The null value, denoted as
  <literal>null</literal>.</para></listitem>

</itemizedlist>

</para>

</simplesect>


<simplesect><title>Lists</title>

<para>Lists are formed by enclosing a whitespace-separated list of
values between square brackets.  For example,

<programlisting>
[ 123 ./foo.nix "abc" (f { x = y; }) ]</programlisting>

defines a list of four elements, the last being the result of a call
to the function <varname>f</varname>.  Note that function calls have
to be enclosed in parentheses.  If they had been omitted, e.g.,

<programlisting>
[ 123 ./foo.nix "abc" f { x = y; } ]</programlisting>

the result would be a list of five elements, the fourth one being a
function and the fifth being a set.</para>

<para>Note that lists are only lazy in values, and they are strict in length.
</para>

</simplesect>


<simplesect><title>Sets</title>

<para>Sets are really the core of the language, since ultimately the
Nix language is all about creating derivations, which are really just
sets of attributes to be passed to build scripts.</para>

<para>Sets are just a list of name/value pairs (called
<emphasis>attributes</emphasis>) enclosed in curly brackets, where
each value is an arbitrary expression terminated by a semicolon.  For
example:

<programlisting>
{ x = 123;
  text = "Hello";
  y = f { bla = 456; };
}</programlisting>

This defines a set with attributes named <varname>x</varname>,
<varname>text</varname>, <varname>y</varname>.  The order of the
attributes is irrelevant.  An attribute name may only occur
once.</para>

<para>Attributes can be selected from a set using the
<literal>.</literal> operator.  For instance,

<programlisting>
{ a = "Foo"; b = "Bar"; }.a</programlisting>

evaluates to <literal>"Foo"</literal>.  It is possible to provide a
default value in an attribute selection using the
<literal>or</literal> keyword.  For example,

<programlisting>
{ a = "Foo"; b = "Bar"; }.c or "Xyzzy"</programlisting>

will evaluate to <literal>"Xyzzy"</literal> because there is no
<varname>c</varname> attribute in the set.</para>

<para>You can use arbitrary double-quoted strings as attribute
names:

<programlisting>
{ "foo ${bar}" = 123; "nix-1.0" = 456; }."foo ${bar}"
</programlisting>

This will evaluate to <literal>123</literal> (Assuming
<literal>bar</literal> is antiquotable). In the case where an
attribute name is just a single antiquotation, the quotes can be
dropped:

<programlisting>
{ foo = 123; }.${bar} or 456 </programlisting>

This will evaluate to <literal>123</literal> if
<literal>bar</literal> evaluates to <literal>"foo"</literal> when
coerced to a string and <literal>456</literal> otherwise (again
assuming <literal>bar</literal> is antiquotable).</para>

<para>In the special case where an attribute name inside of a set declaration
evaluates to <literal>null</literal> (which is normally an error, as
<literal>null</literal> is not antiquotable), that attribute is simply not
added to the set:

<programlisting>
{ ${if foo then "bar" else null} = true; }</programlisting>

This will evaluate to <literal>{}</literal> if <literal>foo</literal>
evaluates to <literal>false</literal>.</para>

<para>A set that has a <literal>__functor</literal> attribute whose value
is callable (i.e. is itself a function or a set with a
<literal>__functor</literal> attribute whose value is callable) can be
applied as if it were a function, with the set itself passed in first
, e.g.,

<programlisting>
let add = { __functor = self: x: x + self.x; };
    inc = add // { x = 1; };
in inc 1
</programlisting>

evaluates to <literal>2</literal>. This can be used to attach metadata to a
function without the caller needing to treat it specially, or to implement
a form of object-oriented programming, for example.

</para>

</simplesect>


</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-constructs">

<title>Language Constructs</title>

<simplesect><title>Recursive sets</title>

<para>Recursive sets are just normal sets, but the attributes can
refer to each other.  For example,

<programlisting>
rec {
  x = y;
  y = 123;
}.x
</programlisting>

evaluates to <literal>123</literal>.  Note that without
<literal>rec</literal> the binding <literal>x = y;</literal> would
refer to the variable <varname>y</varname> in the surrounding scope,
if one exists, and would be invalid if no such variable exists.  That
is, in a normal (non-recursive) set, attributes are not added to the
lexical scope; in a recursive set, they are.</para>

<para>Recursive sets of course introduce the danger of infinite
recursion.  For example,

<programlisting>
rec {
  x = y;
  y = x;
}.x</programlisting>

does not terminate<footnote><para>Actually, Nix detects infinite
recursion in this case and aborts (<quote>infinite recursion
encountered</quote>).</para></footnote>.</para>

</simplesect>


<simplesect xml:id="sect-let-expressions"><title>Let-expressions</title>

<para>A let-expression allows you to define local variables for an
expression.  For instance,

<programlisting>
let
  x = "foo";
  y = "bar";
in x + y</programlisting>

evaluates to <literal>"foobar"</literal>.

</para>

</simplesect>


<simplesect><title>Inheriting attributes</title>

<para>When defining a set or in a let-expression it is often convenient to copy variables
from the surrounding lexical scope (e.g., when you want to propagate
attributes).  This can be shortened using the
<literal>inherit</literal> keyword.  For instance,

<programlisting>
let x = 123; in
{ inherit x;
  y = 456;
}</programlisting>

is equivalent to

<programlisting>
let x = 123; in
{ x = x;
  y = 456;
}</programlisting>

and both evaluate to <literal>{ x = 123; y = 456; }</literal>. (Note that
this works because <varname>x</varname> is added to the lexical scope
by the <literal>let</literal> construct.)  It is also possible to
inherit attributes from another set.  For instance, in this fragment
from <filename>all-packages.nix</filename>,

<programlisting>
  graphviz = (import ../tools/graphics/graphviz) {
    inherit fetchurl stdenv libpng libjpeg expat x11 yacc;
    inherit (xlibs) libXaw;
  };

  xlibs = {
    libX11 = ...;
    libXaw = ...;
    ...
  }

  libpng = ...;
  libjpg = ...;
  ...</programlisting>

the set used in the function call to the function defined in
<filename>../tools/graphics/graphviz</filename> inherits a number of
variables from the surrounding scope (<varname>fetchurl</varname>
... <varname>yacc</varname>), but also inherits
<varname>libXaw</varname> (the X Athena Widgets) from the
<varname>xlibs</varname> (X11 client-side libraries) set.</para>

<para>
Summarizing the fragment

<programlisting>
...
inherit x y z;
inherit (src-set) a b c;
...</programlisting>

is equivalent to

<programlisting>
...
x = x; y = y; z = z;
a = src-set.a; b = src-set.b; c = src-set.c;
...</programlisting>

when used while defining local variables in a let-expression or
while defining a set.</para>

</simplesect>


<simplesect xml:id="ss-functions"><title>Functions</title>

<para>Functions have the following form:

<programlisting>
<replaceable>pattern</replaceable>: <replaceable>body</replaceable></programlisting>

The pattern specifies what the argument of the function must look
like, and binds variables in the body to (parts of) the
argument.  There are three kinds of patterns:</para>

<itemizedlist>


  <listitem><para>If a pattern is a single identifier, then the
  function matches any argument.  Example:

  <programlisting>
let negate = x: !x;
    concat = x: y: x + y;
in if negate true then concat "foo" "bar" else ""</programlisting>

  Note that <function>concat</function> is a function that takes one
  argument and returns a function that takes another argument.  This
  allows partial parameterisation (i.e., only filling some of the
  arguments of a function); e.g.,

  <programlisting>
map (concat "foo") [ "bar" "bla" "abc" ]</programlisting>

  evaluates to <literal>[ "foobar" "foobla"
  "fooabc" ]</literal>.</para></listitem>


  <listitem><para>A <emphasis>set pattern</emphasis> of the form
  <literal>{ name1, name2, &#x2026;, nameN }</literal> matches a set
  containing the listed attributes, and binds the values of those
  attributes to variables in the function body.  For example, the
  function

<programlisting>
{ x, y, z }: z + y + x</programlisting>

  can only be called with a set containing exactly the attributes
  <varname>x</varname>, <varname>y</varname> and
  <varname>z</varname>.  No other attributes are allowed.  If you want
  to allow additional arguments, you can use an ellipsis
  (<literal>...</literal>):

<programlisting>
{ x, y, z, ... }: z + y + x</programlisting>

  This works on any set that contains at least the three named
  attributes.</para>

  <para>It is possible to provide <emphasis>default values</emphasis>
  for attributes, in which case they are allowed to be missing.  A
  default value is specified by writing
  <literal><replaceable>name</replaceable> ?
  <replaceable>e</replaceable></literal>, where
  <replaceable>e</replaceable> is an arbitrary expression.  For example,

<programlisting>
{ x, y ? "foo", z ? "bar" }: z + y + x</programlisting>

  specifies a function that only requires an attribute named
  <varname>x</varname>, but optionally accepts <varname>y</varname>
  and <varname>z</varname>.</para></listitem>


  <listitem><para>An <literal>@</literal>-pattern provides a means of referring
  to the whole value being matched:

<programlisting> args@{ x, y, z, ... }: z + y + x + args.a</programlisting>

but can also be written as:

<programlisting> { x, y, z, ... } @ args: z + y + x + args.a</programlisting>

  Here <varname>args</varname> is bound to the entire argument, which
  is further matched against the pattern <literal>{ x, y, z,
  ... }</literal>. <literal>@</literal>-pattern makes mainly sense with an 
  ellipsis(<literal>...</literal>) as you can access attribute names as 
  <literal>a</literal>, using <literal>args.a</literal>, which was given as an
  additional attribute to the function.
  </para>

  <warning>
   <para>
    The <literal>args@</literal> expression is bound to the argument passed to the function which
    means that attributes with defaults that aren't explicitly specified in the function call
    won't cause an evaluation error, but won't exist in <literal>args</literal>.
   </para>
   <para>
    For instance
<programlisting>
let
  function = args@{ a ? 23, ... }: args;
in
 function {}
</programlisting>
    will evaluate to an empty attribute set.
   </para>
  </warning></listitem>

</itemizedlist>

<para>Note that functions do not have names.  If you want to give them
a name, you can bind them to an attribute, e.g.,

<programlisting>
let concat = { x, y }: x + y;
in concat { x = "foo"; y = "bar"; }</programlisting>

</para>

</simplesect>


<simplesect><title>Conditionals</title>

<para>Conditionals look like this:

<programlisting>
if <replaceable>e1</replaceable> then <replaceable>e2</replaceable> else <replaceable>e3</replaceable></programlisting>

where <replaceable>e1</replaceable> is an expression that should
evaluate to a Boolean value (<literal>true</literal> or
<literal>false</literal>).</para>

</simplesect>


<simplesect><title>Assertions</title>

<para>Assertions are generally used to check that certain requirements
on or between features and dependencies hold.  They look like this:

<programlisting>
assert <replaceable>e1</replaceable>; <replaceable>e2</replaceable></programlisting>

where <replaceable>e1</replaceable> is an expression that should
evaluate to a Boolean value.  If it evaluates to
<literal>true</literal>, <replaceable>e2</replaceable> is returned;
otherwise expression evaluation is aborted and a backtrace is printed.</para>

<example xml:id="ex-subversion-nix"><title>Nix expression for Subversion</title>
<programlisting>
{ localServer ? false
, httpServer ? false
, sslSupport ? false
, pythonBindings ? false
, javaSwigBindings ? false
, javahlBindings ? false
, stdenv, fetchurl
, openssl ? null, httpd ? null, db4 ? null, expat, swig ? null, j2sdk ? null
}:

assert localServer -&gt; db4 != null; <co xml:id="ex-subversion-nix-co-1"/>
assert httpServer -&gt; httpd != null &amp;&amp; httpd.expat == expat; <co xml:id="ex-subversion-nix-co-2"/>
assert sslSupport -&gt; openssl != null &amp;&amp; (httpServer -&gt; httpd.openssl == openssl); <co xml:id="ex-subversion-nix-co-3"/>
assert pythonBindings -&gt; swig != null &amp;&amp; swig.pythonSupport;
assert javaSwigBindings -&gt; swig != null &amp;&amp; swig.javaSupport;
assert javahlBindings -&gt; j2sdk != null;

stdenv.mkDerivation {
  name = "subversion-1.1.1";
  ...
  openssl = if sslSupport then openssl else null; <co xml:id="ex-subversion-nix-co-4"/>
  ...
}</programlisting>
</example>

<para><xref linkend="ex-subversion-nix"/> show how assertions are
used in the Nix expression for Subversion.</para>

<calloutlist>

  <callout arearefs="ex-subversion-nix-co-1">
    <para>This assertion states that if Subversion is to have support
    for local repositories, then Berkeley DB is needed.  So if the
    Subversion function is called with the
    <varname>localServer</varname> argument set to
    <literal>true</literal> but the <varname>db4</varname> argument
    set to <literal>null</literal>, then the evaluation fails.</para>
  </callout>

  <callout arearefs="ex-subversion-nix-co-2">
    <para>This is a more subtle condition: if Subversion is built with
    Apache (<literal>httpServer</literal>) support, then the Expat
    library (an XML library) used by Subversion should be same as the
    one used by Apache.  This is because in this configuration
    Subversion code ends up being linked with Apache code, and if the
    Expat libraries do not match, a build- or runtime link error or
    incompatibility might occur.</para>
  </callout>

  <callout arearefs="ex-subversion-nix-co-3">
    <para>This assertion says that in order for Subversion to have SSL
    support (so that it can access <literal>https</literal> URLs), an
    OpenSSL library must be passed.  Additionally, it says that
    <emphasis>if</emphasis> Apache support is enabled, then Apache's
    OpenSSL should match Subversion's.  (Note that if Apache support
    is not enabled, we don't care about Apache's OpenSSL.)</para>
  </callout>

  <callout arearefs="ex-subversion-nix-co-4">
    <para>The conditional here is not really related to assertions,
    but is worth pointing out: it ensures that if SSL support is
    disabled, then the Subversion derivation is not dependent on
    OpenSSL, even if a non-<literal>null</literal> value was passed.
    This prevents an unnecessary rebuild of Subversion if OpenSSL
    changes.</para>
  </callout>

</calloutlist>

</simplesect>



<simplesect><title>With-expressions</title>

<para>A <emphasis>with-expression</emphasis>,

<programlisting>
with <replaceable>e1</replaceable>; <replaceable>e2</replaceable></programlisting>

introduces the set <replaceable>e1</replaceable> into the lexical
scope of the expression <replaceable>e2</replaceable>.  For instance,

<programlisting>
let as = { x = "foo"; y = "bar"; };
in with as; x + y</programlisting>

evaluates to <literal>"foobar"</literal> since the
<literal>with</literal> adds the <varname>x</varname> and
<varname>y</varname> attributes of <varname>as</varname> to the
lexical scope in the expression <literal>x + y</literal>.  The most
common use of <literal>with</literal> is in conjunction with the
<function>import</function> function.  E.g.,

<programlisting>
with (import ./definitions.nix); ...</programlisting>

makes all attributes defined in the file
<filename>definitions.nix</filename> available as if they were defined
locally in a <literal>let</literal>-expression.</para>

<para>The bindings introduced by <literal>with</literal> do not shadow bindings
introduced by other means, e.g.

<programlisting>
let a = 3; in with { a = 1; }; let a = 4; in with { a = 2; }; ...</programlisting>

establishes the same scope as

<programlisting>
let a = 1; in let a = 2; in let a = 3; in let a = 4; in ...</programlisting>

</para>

</simplesect>


<simplesect><title>Comments</title>

<para>Comments can be single-line, started with a <literal>#</literal>
character, or inline/multi-line, enclosed within <literal>/*
... */</literal>.</para>

</simplesect>


</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-language-operators">

<title>Operators</title>

<para><xref linkend="table-operators"/> lists the operators in the
Nix expression language, in order of precedence (from strongest to
weakest binding).</para>

<table xml:id="table-operators">
  <title>Operators</title>
  <tgroup cols="3">
    <thead>
      <row>
        <entry>Name</entry>
        <entry>Syntax</entry>
        <entry>Associativity</entry>
        <entry>Description</entry>
        <entry>Precedence</entry>
      </row>
    </thead>
    <tbody>
      <row>
        <entry>Select</entry>
        <entry><replaceable>e</replaceable> <literal>.</literal>
        <replaceable>attrpath</replaceable>
        [ <literal>or</literal> <replaceable>def</replaceable> ]
        </entry>
        <entry>none</entry>
        <entry>Select attribute denoted by the attribute path
        <replaceable>attrpath</replaceable> from set
        <replaceable>e</replaceable>.  (An attribute path is a
        dot-separated list of attribute names.)  If the attribute
        doesn&#x2019;t exist, return <replaceable>def</replaceable> if
        provided, otherwise abort evaluation.</entry>
        <entry>1</entry>
      </row>
      <row>
        <entry>Application</entry>
        <entry><replaceable>e1</replaceable> <replaceable>e2</replaceable></entry>
        <entry>left</entry>
        <entry>Call function <replaceable>e1</replaceable> with
        argument <replaceable>e2</replaceable>.</entry>
        <entry>2</entry>
      </row>
      <row>
        <entry>Arithmetic Negation</entry>
        <entry><literal>-</literal> <replaceable>e</replaceable></entry>
        <entry>none</entry>
        <entry>Arithmetic negation.</entry>
        <entry>3</entry>
      </row>
      <row>
        <entry>Has Attribute</entry>
        <entry><replaceable>e</replaceable> <literal>?</literal>
        <replaceable>attrpath</replaceable></entry>
        <entry>none</entry>
        <entry>Test whether set <replaceable>e</replaceable> contains
        the attribute denoted by <replaceable>attrpath</replaceable>;
        return <literal>true</literal> or
        <literal>false</literal>.</entry>
        <entry>4</entry>
      </row>
      <row>
        <entry>List Concatenation</entry>
        <entry><replaceable>e1</replaceable> <literal>++</literal> <replaceable>e2</replaceable></entry>
        <entry>right</entry>
        <entry>List concatenation.</entry>
        <entry>5</entry>
      </row>
      <row>
        <entry>Multiplication</entry>
        <entry>
          <replaceable>e1</replaceable> <literal>*</literal> <replaceable>e2</replaceable>,
        </entry>
        <entry>left</entry>
        <entry>Arithmetic multiplication.</entry>
        <entry>6</entry>
      </row>
      <row>
        <entry>Division</entry>
        <entry>
          <replaceable>e1</replaceable> <literal>/</literal> <replaceable>e2</replaceable>
        </entry>
        <entry>left</entry>
        <entry>Arithmetic division.</entry>
        <entry>6</entry>
      </row>
      <row>
        <entry>Addition</entry>
        <entry>
          <replaceable>e1</replaceable> <literal>+</literal> <replaceable>e2</replaceable>
        </entry>
        <entry>left</entry>
        <entry>Arithmetic addition.</entry>
        <entry>7</entry>
      </row>
      <row>
        <entry>Subtraction</entry>
        <entry>
          <replaceable>e1</replaceable> <literal>-</literal> <replaceable>e2</replaceable>
        </entry>
        <entry>left</entry>
        <entry>Arithmetic subtraction.</entry>
        <entry>7</entry>
      </row>
      <row>
        <entry>String Concatenation</entry>
        <entry>
          <replaceable>string1</replaceable> <literal>+</literal> <replaceable>string2</replaceable>
        </entry>
        <entry>left</entry>
        <entry>String concatenation.</entry>
        <entry>7</entry>
      </row>
      <row>
        <entry>Not</entry>
        <entry><literal>!</literal> <replaceable>e</replaceable></entry>
        <entry>none</entry>
        <entry>Boolean negation.</entry>
        <entry>8</entry>
      </row>
      <row>
        <entry>Update</entry>
        <entry><replaceable>e1</replaceable> <literal>//</literal>
        <replaceable>e2</replaceable></entry>
        <entry>right</entry>
        <entry>Return a set consisting of the attributes in
        <replaceable>e1</replaceable> and
        <replaceable>e2</replaceable> (with the latter taking
        precedence over the former in case of equally named
        attributes).</entry>
        <entry>9</entry>
      </row>
      <row>
        <entry>Less Than</entry>
        <entry>
          <replaceable>e1</replaceable> <literal>&lt;</literal> <replaceable>e2</replaceable>,
        </entry>
        <entry>none</entry>
        <entry>Arithmetic comparison.</entry>
        <entry>10</entry>
      </row>
      <row>
        <entry>Less Than or Equal To</entry>
        <entry>
          <replaceable>e1</replaceable> <literal>&lt;=</literal> <replaceable>e2</replaceable>
        </entry>
        <entry>none</entry>
        <entry>Arithmetic comparison.</entry>
        <entry>10</entry>
      </row>
      <row>
        <entry>Greater Than</entry>
        <entry>
          <replaceable>e1</replaceable> <literal>&gt;</literal> <replaceable>e2</replaceable>
        </entry>
        <entry>none</entry>
        <entry>Arithmetic comparison.</entry>
        <entry>10</entry>
      </row>
      <row>
        <entry>Greater Than or Equal To</entry>
        <entry>
          <replaceable>e1</replaceable> <literal>&gt;=</literal> <replaceable>e2</replaceable>
        </entry>
        <entry>none</entry>
        <entry>Arithmetic comparison.</entry>
        <entry>10</entry>
      </row>
      <row>
        <entry>Equality</entry>
        <entry>
          <replaceable>e1</replaceable> <literal>==</literal> <replaceable>e2</replaceable>
        </entry>
        <entry>none</entry>
        <entry>Equality.</entry>
        <entry>11</entry>
      </row>
      <row>
        <entry>Inequality</entry>
        <entry>
          <replaceable>e1</replaceable> <literal>!=</literal> <replaceable>e2</replaceable>
        </entry>
        <entry>none</entry>
        <entry>Inequality.</entry>
        <entry>11</entry>
      </row>
      <row>
        <entry>Logical AND</entry>
        <entry><replaceable>e1</replaceable> <literal>&amp;&amp;</literal>
        <replaceable>e2</replaceable></entry>
        <entry>left</entry>
        <entry>Logical AND.</entry>
        <entry>12</entry>
      </row>
      <row>
        <entry>Logical OR</entry>
        <entry><replaceable>e1</replaceable> <literal>||</literal>
        <replaceable>e2</replaceable></entry>
        <entry>left</entry>
        <entry>Logical OR.</entry>
        <entry>13</entry>
      </row>
      <row>
        <entry>Logical Implication</entry>
        <entry><replaceable>e1</replaceable> <literal>-&gt;</literal>
        <replaceable>e2</replaceable></entry>
        <entry>none</entry>
        <entry>Logical implication (equivalent to
        <literal>!<replaceable>e1</replaceable> ||
        <replaceable>e2</replaceable></literal>).</entry>
        <entry>14</entry>
      </row>
    </tbody>
  </tgroup>
</table>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-derivation">

<title>Derivations</title>

<para>The most important built-in function is
<function>derivation</function>, which is used to describe a single
derivation (a build action).  It takes as input a set, the attributes
of which specify the inputs of the build.</para>

<itemizedlist>

  <listitem xml:id="attr-system"><para>There must be an attribute named
  <varname>system</varname> whose value must be a string specifying a
  Nix platform identifier, such as <literal>"i686-linux"</literal> or
  <literal>"x86_64-darwin"</literal><footnote><para>To figure out
  your platform identifier, look at the line <quote>Checking for the
  canonical Nix system name</quote> in the output of Nix's
  <filename>configure</filename> script.</para></footnote> The build
  can only be performed on a machine and operating system matching the
  platform identifier.  (Nix can automatically forward builds for
  other platforms by forwarding them to other machines; see <xref linkend="chap-distributed-builds"/>.)</para></listitem>

  <listitem><para>There must be an attribute named
  <varname>name</varname> whose value must be a string.  This is used
  as a symbolic name for the package by <command>nix-env</command>,
  and it is appended to the output paths of the
  derivation.</para></listitem>

  <listitem><para>There must be an attribute named
  <varname>builder</varname> that identifies the program that is
  executed to perform the build.  It can be either a derivation or a
  source (a local file reference, e.g.,
  <filename>./builder.sh</filename>).</para></listitem>

  <listitem><para>Every attribute is passed as an environment variable
  to the builder.  Attribute values are translated to environment
  variables as follows:

    <itemizedlist>

      <listitem><para>Strings and numbers are just passed
      verbatim.</para></listitem>

      <listitem><para>A <emphasis>path</emphasis> (e.g.,
      <filename>../foo/sources.tar</filename>) causes the referenced
      file to be copied to the store; its location in the store is put
      in the environment variable.  The idea is that all sources
      should reside in the Nix store, since all inputs to a derivation
      should reside in the Nix store.</para></listitem>

      <listitem><para>A <emphasis>derivation</emphasis> causes that
      derivation to be built prior to the present derivation; its
      default output path is put in the environment
      variable.</para></listitem>

      <listitem><para>Lists of the previous types are also allowed.
      They are simply concatenated, separated by
      spaces.</para></listitem>

      <listitem><para><literal>true</literal> is passed as the string
      <literal>1</literal>, <literal>false</literal> and
      <literal>null</literal> are passed as an empty string.
      </para></listitem>
    </itemizedlist>

  </para></listitem>

  <listitem><para>The optional attribute <varname>args</varname>
  specifies command-line arguments to be passed to the builder.  It
  should be a list.</para></listitem>

  <listitem><para>The optional attribute <varname>outputs</varname>
  specifies a list of symbolic outputs of the derivation.  By default,
  a derivation produces a single output path, denoted as
  <literal>out</literal>.  However, derivations can produce multiple
  output paths.  This is useful because it allows outputs to be
  downloaded or garbage-collected separately.  For instance, imagine a
  library package that provides a dynamic library, header files, and
  documentation.  A program that links against the library doesn&#x2019;t
  need the header files and documentation at runtime, and it doesn&#x2019;t
  need the documentation at build time.  Thus, the library package
  could specify:
<programlisting>
outputs = [ "lib" "headers" "doc" ];
</programlisting>
  This will cause Nix to pass environment variables
  <literal>lib</literal>, <literal>headers</literal> and
  <literal>doc</literal> to the builder containing the intended store
  paths of each output.  The builder would typically do something like
<programlisting>
./configure --libdir=$lib/lib --includedir=$headers/include --docdir=$doc/share/doc
</programlisting>
  for an Autoconf-style package.  You can refer to each output of a
  derivation by selecting it as an attribute, e.g.
<programlisting>
buildInputs = [ pkg.lib pkg.headers ];
</programlisting>
  The first element of <varname>outputs</varname> determines the
  <emphasis>default output</emphasis>.  Thus, you could also write
<programlisting>
buildInputs = [ pkg pkg.headers ];
</programlisting>
  since <literal>pkg</literal> is equivalent to
  <literal>pkg.lib</literal>.</para></listitem>

</itemizedlist>

<para>The function <function>mkDerivation</function> in the Nixpkgs
standard environment is a wrapper around
<function>derivation</function> that adds a default value for
<varname>system</varname> and always uses Bash as the builder, to
which the supplied builder is passed as a command-line argument.  See
the Nixpkgs manual for details.</para>

<para>The builder is executed as follows:

<itemizedlist>

  <listitem><para>A temporary directory is created under the directory
  specified by <envar>TMPDIR</envar> (default
  <filename>/tmp</filename>) where the build will take place.  The
  current directory is changed to this directory.</para></listitem>

  <listitem><para>The environment is cleared and set to the derivation
  attributes, as specified above.</para></listitem>

  <listitem><para>In addition, the following variables are set:

  <itemizedlist>

    <listitem><para><envar>NIX_BUILD_TOP</envar> contains the path of
    the temporary directory for this build.</para></listitem>

    <listitem><para>Also, <envar>TMPDIR</envar>,
    <envar>TEMPDIR</envar>, <envar>TMP</envar>, <envar>TEMP</envar>
    are set to point to the temporary directory.  This is to prevent
    the builder from accidentally writing temporary files anywhere
    else.  Doing so might cause interference by other
    processes.</para></listitem>

    <listitem><para><envar>PATH</envar> is set to
    <filename>/path-not-set</filename> to prevent shells from
    initialising it to their built-in default value.</para></listitem>

    <listitem><para><envar>HOME</envar> is set to
    <filename>/homeless-shelter</filename> to prevent programs from
    using <filename>/etc/passwd</filename> or the like to find the
    user's home directory, which could cause impurity.  Usually, when
    <envar>HOME</envar> is set, it is used as the location of the home
    directory, even if it points to a non-existent
    path.</para></listitem>

    <listitem><para><envar>NIX_STORE</envar> is set to the path of the
    top-level Nix store directory (typically,
    <filename>/nix/store</filename>).</para></listitem>

    <listitem><para>For each output declared in
    <varname>outputs</varname>, the corresponding environment variable
    is set to point to the intended path in the Nix store for that
    output.  Each output path is a concatenation of the cryptographic
    hash of all build inputs, the <varname>name</varname> attribute
    and the output name.  (The output name is omitted if it&#x2019;s
    <literal>out</literal>.)</para></listitem>

  </itemizedlist>

  </para></listitem>

  <listitem><para>If an output path already exists, it is removed.
  Also, locks are acquired to prevent multiple Nix instances from
  performing the same build at the same time.</para></listitem>

  <listitem><para>A log of the combined standard output and error is
  written to <filename>/nix/var/log/nix</filename>.</para></listitem>

  <listitem><para>The builder is executed with the arguments specified
  by the attribute <varname>args</varname>.  If it exits with exit
  code 0, it is considered to have succeeded.</para></listitem>

  <listitem><para>The temporary directory is removed (unless the
  <option>-K</option> option was specified).</para></listitem>

  <listitem><para>If the build was successful, Nix scans each output
  path for references to input paths by looking for the hash parts of
  the input paths.  Since these are potential runtime dependencies,
  Nix registers them as dependencies of the output
  paths.</para></listitem>

  <listitem><para>After the build, Nix sets the last-modified
  timestamp on all files in the build result to 1 (00:00:01 1/1/1970
  UTC), sets the group to the default group, and sets the mode of the
  file to 0444 or 0555 (i.e., read-only, with execute permission
  enabled if the file was originally executable).  Note that possible
  <literal>setuid</literal> and <literal>setgid</literal> bits are
  cleared.  Setuid and setgid programs are not currently supported by
  Nix.  This is because the Nix archives used in deployment have no
  concept of ownership information, and because it makes the build
  result dependent on the user performing the build.</para></listitem>

</itemizedlist>

</para>

<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-advanced-attributes">

<title>Advanced Attributes</title>

<para>Derivations can declare some infrequently used optional
attributes.</para>

<variablelist>

  <varlistentry xml:id="adv-attr-allowedReferences"><term><varname>allowedReferences</varname></term>

    <listitem><para>The optional attribute
    <varname>allowedReferences</varname> specifies a list of legal
    references (dependencies) of the output of the builder.  For
    example,

<programlisting>
allowedReferences = [];
</programlisting>

    enforces that the output of a derivation cannot have any runtime
    dependencies on its inputs.  To allow an output to have a runtime
    dependency on itself, use <literal>"out"</literal> as a list item.
    This is used in NixOS to check that generated files such as
    initial ramdisks for booting Linux don&#x2019;t have accidental
    dependencies on other paths in the Nix store.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="adv-attr-allowedRequisites"><term><varname>allowedRequisites</varname></term>

    <listitem><para>This attribute is similar to
    <varname>allowedReferences</varname>, but it specifies the legal
    requisites of the whole closure, so all the dependencies
    recursively.  For example,

<programlisting>
allowedRequisites = [ foobar ];
</programlisting>

    enforces that the output of a derivation cannot have any other
    runtime dependency than <varname>foobar</varname>, and in addition
    it enforces that <varname>foobar</varname> itself doesn't
    introduce any other dependency itself.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="adv-attr-disallowedReferences"><term><varname>disallowedReferences</varname></term>

    <listitem><para>The optional attribute
    <varname>disallowedReferences</varname> specifies a list of illegal
    references (dependencies) of the output of the builder.  For
    example,

<programlisting>
disallowedReferences = [ foo ];
</programlisting>

    enforces that the output of a derivation cannot have a direct runtime
    dependencies on the derivation <varname>foo</varname>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="adv-attr-disallowedRequisites"><term><varname>disallowedRequisites</varname></term>

    <listitem><para>This attribute is similar to
    <varname>disallowedReferences</varname>, but it specifies illegal
    requisites for the whole closure, so all the dependencies
    recursively.  For example,

<programlisting>
disallowedRequisites = [ foobar ];
</programlisting>

    enforces that the output of a derivation cannot have any
    runtime dependency on <varname>foobar</varname> or any other derivation
    depending recursively on <varname>foobar</varname>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="adv-attr-exportReferencesGraph"><term><varname>exportReferencesGraph</varname></term>

    <listitem><para>This attribute allows builders access to the
    references graph of their inputs.  The attribute is a list of
    inputs in the Nix store whose references graph the builder needs
    to know.  The value of this attribute should be a list of pairs
    <literal>[ <replaceable>name1</replaceable>
    <replaceable>path1</replaceable> <replaceable>name2</replaceable>
    <replaceable>path2</replaceable> <replaceable>...</replaceable>
    ]</literal>.  The references graph of each
    <replaceable>pathN</replaceable> will be stored in a text file
    <replaceable>nameN</replaceable> in the temporary build directory.
    The text files have the format used by <command>nix-store
    --register-validity</command> (with the deriver fields left
    empty).  For example, when the following derivation is built:

<programlisting>
derivation {
  ...
  exportReferencesGraph = [ "libfoo-graph" libfoo ];
};
</programlisting>

    the references graph of <literal>libfoo</literal> is placed in the
    file <filename>libfoo-graph</filename> in the temporary build
    directory.</para>

    <para><varname>exportReferencesGraph</varname> is useful for
    builders that want to do something with the closure of a store
    path.  Examples include the builders in NixOS that generate the
    initial ramdisk for booting Linux (a <command>cpio</command>
    archive containing the closure of the boot script) and the
    ISO-9660 image for the installation CD (which is populated with a
    Nix store containing the closure of a bootable NixOS
    configuration).</para></listitem>

  </varlistentry>


  <varlistentry xml:id="adv-attr-impureEnvVars"><term><varname>impureEnvVars</varname></term>

    <listitem><para>This attribute allows you to specify a list of
    environment variables that should be passed from the environment
    of the calling user to the builder.  Usually, the environment is
    cleared completely when the builder is executed, but with this
    attribute you can allow specific environment variables to be
    passed unmodified.  For example, <function>fetchurl</function> in
    Nixpkgs has the line

<programlisting>
impureEnvVars = [ "http_proxy" "https_proxy" <replaceable>...</replaceable> ];
</programlisting>

    to make it use the proxy server configuration specified by the
    user in the environment variables <envar>http_proxy</envar> and
    friends.</para>

    <para>This attribute is only allowed in <link linkend="fixed-output-drvs">fixed-output derivations</link>, where
    impurities such as these are okay since (the hash of) the output
    is known in advance.  It is ignored for all other
    derivations.</para>

    <warning><para><varname>impureEnvVars</varname> implementation takes
    environment variables from the current builder process. When a daemon is
    building its environmental variables are used. Without the daemon, the
    environmental variables come from the environment of the
    <command>nix-build</command>.</para></warning></listitem>

  </varlistentry>


  <varlistentry xml:id="fixed-output-drvs">
    <term xml:id="adv-attr-outputHash"><varname>outputHash</varname></term>
    <term xml:id="adv-attr-outputHashAlgo"><varname>outputHashAlgo</varname></term>
    <term xml:id="adv-attr-outputHashMode"><varname>outputHashMode</varname></term>

    <listitem><para>These attributes declare that the derivation is a
    so-called <emphasis>fixed-output derivation</emphasis>, which
    means that a cryptographic hash of the output is already known in
    advance.  When the build of a fixed-output derivation finishes,
    Nix computes the cryptographic hash of the output and compares it
    to the hash declared with these attributes.  If there is a
    mismatch, the build fails.</para>

    <para>The rationale for fixed-output derivations is derivations
    such as those produced by the <function>fetchurl</function>
    function.  This function downloads a file from a given URL.  To
    ensure that the downloaded file has not been modified, the caller
    must also specify a cryptographic hash of the file.  For example,

<programlisting>
fetchurl {
  url = "http://ftp.gnu.org/pub/gnu/hello/hello-2.1.1.tar.gz";
  sha256 = "1md7jsfd8pa45z73bz1kszpp01yw6x5ljkjk2hx7wl800any6465";
}
</programlisting>

    It sometimes happens that the URL of the file changes, e.g.,
    because servers are reorganised or no longer available.  We then
    must update the call to <function>fetchurl</function>, e.g.,

<programlisting>
fetchurl {
  url = "ftp://ftp.nluug.nl/pub/gnu/hello/hello-2.1.1.tar.gz";
  sha256 = "1md7jsfd8pa45z73bz1kszpp01yw6x5ljkjk2hx7wl800any6465";
}
</programlisting>

    If a <function>fetchurl</function> derivation was treated like a
    normal derivation, the output paths of the derivation and
    <emphasis>all derivations depending on it</emphasis> would change.
    For instance, if we were to change the URL of the Glibc source
    distribution in Nixpkgs (a package on which almost all other
    packages depend) massive rebuilds would be needed.  This is
    unfortunate for a change which we know cannot have a real effect
    as it propagates upwards through the dependency graph.</para>

    <para>For fixed-output derivations, on the other hand, the name of
    the output path only depends on the <varname>outputHash*</varname>
    and <varname>name</varname> attributes, while all other attributes
    are ignored for the purpose of computing the output path.  (The
    <varname>name</varname> attribute is included because it is part
    of the path.)</para>

    <para>As an example, here is the (simplified) Nix expression for
    <varname>fetchurl</varname>:

<programlisting>
{ stdenv, curl }: # The <command>curl</command> program is used for downloading.

{ url, sha256 }:

stdenv.mkDerivation {
  name = baseNameOf (toString url);
  builder = ./builder.sh;
  buildInputs = [ curl ];

  # This is a fixed-output derivation; the output must be a regular
  # file with SHA256 hash <varname>sha256</varname>.
  outputHashMode = "flat";
  outputHashAlgo = "sha256";
  outputHash = sha256;

  inherit url;
}
</programlisting>

    </para>

    <para>The <varname>outputHashAlgo</varname> attribute specifies
    the hash algorithm used to compute the hash.  It can currently be
    <literal>"sha1"</literal>, <literal>"sha256"</literal> or
    <literal>"sha512"</literal>.</para>

    <para>The <varname>outputHashMode</varname> attribute determines
    how the hash is computed.  It must be one of the following two
    values:

    <variablelist>

      <varlistentry><term><literal>"flat"</literal></term>

        <listitem><para>The output must be a non-executable regular
        file.  If it isn&#x2019;t, the build fails.  The hash is simply
        computed over the contents of that file (so it&#x2019;s equal to what
        Unix commands like <command>sha256sum</command> or
        <command>sha1sum</command> produce).</para>

        <para>This is the default.</para></listitem>

      </varlistentry>

      <varlistentry><term><literal>"recursive"</literal></term>

        <listitem><para>The hash is computed over the NAR archive dump
        of the output (i.e., the result of <link linkend="refsec-nix-store-dump"><command>nix-store
        --dump</command></link>).  In this case, the output can be
        anything, including a directory tree.</para></listitem>

      </varlistentry>

    </variablelist>

    </para>

    <para>The <varname>outputHash</varname> attribute, finally, must
    be a string containing the hash in either hexadecimal or base-32
    notation.  (See the <link linkend="sec-nix-hash"><command>nix-hash</command> command</link>
    for information about converting to and from base-32
    notation.)</para></listitem>

  </varlistentry>


  <varlistentry xml:id="adv-attr-passAsFile"><term><varname>passAsFile</varname></term>

    <listitem><para>A list of names of attributes that should be
    passed via files rather than environment variables.  For example,
    if you have

    <programlisting>
passAsFile = ["big"];
big = "a very long string";
    </programlisting>

    then when the builder runs, the environment variable
    <envar>bigPath</envar> will contain the absolute path to a
    temporary file containing <literal>a very long
    string</literal>. That is, for any attribute
    <replaceable>x</replaceable> listed in
    <varname>passAsFile</varname>, Nix will pass an environment
    variable <envar><replaceable>x</replaceable>Path</envar> holding
    the path of the file containing the value of attribute
    <replaceable>x</replaceable>. This is useful when you need to pass
    large strings to a builder, since most operating systems impose a
    limit on the size of the environment (typically, a few hundred
    kilobyte).</para></listitem>

  </varlistentry>


  <varlistentry xml:id="adv-attr-preferLocalBuild"><term><varname>preferLocalBuild</varname></term>

    <listitem><para>If this attribute is set to
    <literal>true</literal> and <link linkend="chap-distributed-builds">distributed building is
    enabled</link>, then, if possible, the derivaton will be built
    locally instead of forwarded to a remote machine.  This is
    appropriate for trivial builders where the cost of doing a
    download or remote build would exceed the cost of building
    locally.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="adv-attr-allowSubstitutes"><term><varname>allowSubstitutes</varname></term>

    <listitem>
    <para>If this attribute is set to
    <literal>false</literal>, then Nix will always build this
    derivation; it will not try to substitute its outputs. This is
    useful for very trivial derivations (such as
    <function>writeText</function> in Nixpkgs) that are cheaper to
    build than to substitute from a binary cache.</para>

    <note><para>You need to have a builder configured which satisfies
    the derivation&#x2019;s <literal>system</literal> attribute, since the
    derivation cannot be substituted. Thus it is usually a good idea
    to align <literal>system</literal> with
    <literal>builtins.currentSystem</literal> when setting
    <literal>allowSubstitutes</literal> to <literal>false</literal>.
    For most trivial derivations this should be the case.
    </para></note>
    </listitem>

  </varlistentry>


</variablelist>

</section>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-builtins">

<title>Built-in Functions</title>

<para>This section lists the functions and constants built into the
Nix expression evaluator.  (The built-in function
<function>derivation</function> is discussed above.)  Some built-ins,
such as <function>derivation</function>, are always in scope of every
Nix expression; you can just access them right away.  But to prevent
polluting the namespace too much, most built-ins are not in scope.
Instead, you can access them through the <varname>builtins</varname>
built-in value, which is a set that contains all built-in functions
and values.  For instance, <function>derivation</function> is also
available as <function>builtins.derivation</function>.</para>


<variablelist>


  <varlistentry xml:id="builtin-abort">
    <term><function>abort</function> <replaceable>s</replaceable></term>
    <term><function>builtins.abort</function> <replaceable>s</replaceable></term>

    <listitem><para>Abort Nix expression evaluation, print error
    message <replaceable>s</replaceable>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-add">
    <term><function>builtins.add</function>
    <replaceable>e1</replaceable> <replaceable>e2</replaceable>
    </term>

    <listitem><para>Return the sum of the numbers
    <replaceable>e1</replaceable> and
    <replaceable>e2</replaceable>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-all">
    <term><function>builtins.all</function>
    <replaceable>pred</replaceable> <replaceable>list</replaceable></term>

    <listitem><para>Return <literal>true</literal> if the function
    <replaceable>pred</replaceable> returns <literal>true</literal>
    for all elements of <replaceable>list</replaceable>,
    and <literal>false</literal> otherwise.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-any">
    <term><function>builtins.any</function>
    <replaceable>pred</replaceable> <replaceable>list</replaceable></term>

    <listitem><para>Return <literal>true</literal> if the function
    <replaceable>pred</replaceable> returns <literal>true</literal>
    for at least one element of <replaceable>list</replaceable>,
    and <literal>false</literal> otherwise.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-attrNames">
    <term><function>builtins.attrNames</function>
    <replaceable>set</replaceable></term>

    <listitem><para>Return the names of the attributes in the set
    <replaceable>set</replaceable> in an alphabetically sorted list.  For instance,
    <literal>builtins.attrNames { y = 1; x = "foo"; }</literal>
    evaluates to <literal>[ "x" "y" ]</literal>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-attrValues">
    <term><function>builtins.attrValues</function>
    <replaceable>set</replaceable></term>

    <listitem><para>Return the values of the attributes in the set
    <replaceable>set</replaceable> in the order corresponding to the
    sorted attribute names.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-baseNameOf">
    <term><function>baseNameOf</function> <replaceable>s</replaceable></term>

    <listitem><para>Return the <emphasis>base name</emphasis> of the
    string <replaceable>s</replaceable>, that is, everything following
    the final slash in the string.  This is similar to the GNU
    <command>basename</command> command.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-bitAnd">
    <term><function>builtins.bitAnd</function>
    <replaceable>e1</replaceable> <replaceable>e2</replaceable></term>

    <listitem><para>Return the bitwise AND of the integers
    <replaceable>e1</replaceable> and
    <replaceable>e2</replaceable>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-bitOr">
    <term><function>builtins.bitOr</function>
    <replaceable>e1</replaceable> <replaceable>e2</replaceable></term>

    <listitem><para>Return the bitwise OR of the integers
    <replaceable>e1</replaceable> and
    <replaceable>e2</replaceable>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-bitXor">
    <term><function>builtins.bitXor</function>
    <replaceable>e1</replaceable> <replaceable>e2</replaceable></term>

    <listitem><para>Return the bitwise XOR of the integers
    <replaceable>e1</replaceable> and
    <replaceable>e2</replaceable>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-builtins">
    <term><varname>builtins</varname></term>

    <listitem><para>The set <varname>builtins</varname> contains all
    the built-in functions and values.  You can use
    <varname>builtins</varname> to test for the availability of
    features in the Nix installation, e.g.,

<programlisting>
if builtins ? getEnv then builtins.getEnv "PATH" else ""</programlisting>

    This allows a Nix expression to fall back gracefully on older Nix
    installations that don&#x2019;t have the desired built-in
    function.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-compareVersions">
    <term><function>builtins.compareVersions</function>
    <replaceable>s1</replaceable> <replaceable>s2</replaceable></term>

    <listitem><para>Compare two strings representing versions and
    return <literal>-1</literal> if version
    <replaceable>s1</replaceable> is older than version
    <replaceable>s2</replaceable>, <literal>0</literal> if they are
    the same, and <literal>1</literal> if
    <replaceable>s1</replaceable> is newer than
    <replaceable>s2</replaceable>.  The version comparison algorithm
    is the same as the one used by <link linkend="ssec-version-comparisons"><command>nix-env
    -u</command></link>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-concatLists">
    <term><function>builtins.concatLists</function>
    <replaceable>lists</replaceable></term>

    <listitem><para>Concatenate a list of lists into a single
    list.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="builtin-concatStringsSep">
    <term><function>builtins.concatStringsSep</function>
    <replaceable>separator</replaceable> <replaceable>list</replaceable></term>

    <listitem><para>Concatenate a list of strings with a separator
    between each element, e.g. <literal>concatStringsSep "/"
    ["usr" "local" "bin"] == "usr/local/bin"</literal></para></listitem>

  </varlistentry>

  <varlistentry xml:id="builtin-currentSystem">
    <term><varname>builtins.currentSystem</varname></term>

    <listitem><para>The built-in value <varname>currentSystem</varname>
    evaluates to the Nix platform identifier for the Nix installation
    on which the expression is being evaluated, such as
    <literal>"i686-linux"</literal> or
    <literal>"x86_64-darwin"</literal>.</para></listitem>

  </varlistentry>


  <!--
  <varlistentry><term><function>currentTime</function></term>

    <listitem><para>The built-in value <varname>currentTime</varname>
    returns the current system time in seconds since 00:00:00 1/1/1970
    UTC.  Due to the evaluation model of Nix expressions
    (<emphasis>maximal laziness</emphasis>), it always yields the same
    value within an execution of Nix.</para></listitem>

  </varlistentry>
  -->


  <!--
  <varlistentry><term><function>dependencyClosure</function></term>

    <listitem><para>TODO</para></listitem>

  </varlistentry>
  -->


  <varlistentry xml:id="builtin-deepSeq">
    <term><function>builtins.deepSeq</function>
    <replaceable>e1</replaceable> <replaceable>e2</replaceable></term>

    <listitem><para>This is like <literal>seq
    <replaceable>e1</replaceable>
    <replaceable>e2</replaceable></literal>, except that
    <replaceable>e1</replaceable> is evaluated
    <emphasis>deeply</emphasis>: if it&#x2019;s a list or set, its elements
    or attributes are also evaluated recursively.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-derivation">
    <term><function>derivation</function>
    <replaceable>attrs</replaceable></term>
    <term><function>builtins.derivation</function>
    <replaceable>attrs</replaceable></term>

    <listitem><para><function>derivation</function> is described in
    <xref linkend="ssec-derivation"/>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-dirOf">
    <term><function>dirOf</function> <replaceable>s</replaceable></term>
    <term><function>builtins.dirOf</function> <replaceable>s</replaceable></term>

    <listitem><para>Return the directory part of the string
    <replaceable>s</replaceable>, that is, everything before the final
    slash in the string.  This is similar to the GNU
    <command>dirname</command> command.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-div">
    <term><function>builtins.div</function>
    <replaceable>e1</replaceable> <replaceable>e2</replaceable></term>

    <listitem><para>Return the quotient of the numbers
    <replaceable>e1</replaceable> and
    <replaceable>e2</replaceable>.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="builtin-elem">
    <term><function>builtins.elem</function>
    <replaceable>x</replaceable> <replaceable>xs</replaceable></term>

    <listitem><para>Return <literal>true</literal> if a value equal to
    <replaceable>x</replaceable> occurs in the list
    <replaceable>xs</replaceable>, and <literal>false</literal>
    otherwise.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-elemAt">
    <term><function>builtins.elemAt</function>
    <replaceable>xs</replaceable> <replaceable>n</replaceable></term>

    <listitem><para>Return element <replaceable>n</replaceable> from
    the list <replaceable>xs</replaceable>.  Elements are counted
    starting from 0.  A fatal error occurs if the index is out of
    bounds.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-fetchurl">
    <term><function>builtins.fetchurl</function>
    <replaceable>url</replaceable></term>

    <listitem><para>Download the specified URL and return the path of
    the downloaded file. This function is not available if <link linkend="conf-restrict-eval">restricted evaluation mode</link> is
    enabled.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-fetchTarball">
    <term><function>fetchTarball</function>
    <replaceable>url</replaceable></term>
    <term><function>builtins.fetchTarball</function>
    <replaceable>url</replaceable></term>

    <listitem><para>Download the specified URL, unpack it and return
    the path of the unpacked tree. The file must be a tape archive
    (<filename>.tar</filename>) compressed with
    <literal>gzip</literal>, <literal>bzip2</literal> or
    <literal>xz</literal>. The top-level path component of the files
    in the tarball is removed, so it is best if the tarball contains a
    single directory at top level. The typical use of the function is
    to obtain external Nix expression dependencies, such as a
    particular version of Nixpkgs, e.g.

<programlisting>
with import (fetchTarball https://github.com/NixOS/nixpkgs/archive/nixos-14.12.tar.gz) {};

stdenv.mkDerivation { &#x2026; }
</programlisting>
    </para>

    <para>The fetched tarball is cached for a certain amount of time
    (1 hour by default) in <filename>~/.cache/nix/tarballs/</filename>.
    You can change the cache timeout either on the command line with
    <option>--option tarball-ttl <replaceable>number of seconds</replaceable></option> or
    in the Nix configuration file with this option:
    <literal><xref linkend="conf-tarball-ttl"/> <replaceable>number of seconds to cache</replaceable></literal>.
    </para>

    <para>Note that when obtaining the hash with <varname>nix-prefetch-url
    </varname> the option <varname>--unpack</varname> is required.
    </para>

    <para>This function can also verify the contents against a hash.
    In that case, the function takes a set instead of a URL. The set
    requires the attribute <varname>url</varname> and the attribute
    <varname>sha256</varname>, e.g.

<programlisting>
with import (fetchTarball {
  url = "https://github.com/NixOS/nixpkgs/archive/nixos-14.12.tar.gz";
  sha256 = "1jppksrfvbk5ypiqdz4cddxdl8z6zyzdb2srq8fcffr327ld5jj2";
}) {};

stdenv.mkDerivation { &#x2026; }
</programlisting>

    </para>

    <para>This function is not available if <link linkend="conf-restrict-eval">restricted evaluation mode</link> is
    enabled.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="builtin-fetchGit">
    <term>
      <function>builtins.fetchGit</function>
      <replaceable>args</replaceable>
    </term>

    <listitem>
      <para>
        Fetch a path from git. <replaceable>args</replaceable> can be
        a URL, in which case the HEAD of the repo at that URL is
        fetched. Otherwise, it can be an attribute with the following
        attributes (all except <varname>url</varname> optional):
      </para>

      <variablelist>
        <varlistentry>
          <term>url</term>
          <listitem>
            <para>
              The URL of the repo.
            </para>
          </listitem>
        </varlistentry>
        <varlistentry>
          <term>name</term>
          <listitem>
            <para>
              The name of the directory the repo should be exported to
              in the store. Defaults to the basename of the URL.
            </para>
          </listitem>
        </varlistentry>
        <varlistentry>
          <term>rev</term>
          <listitem>
            <para>
              The git revision to fetch. Defaults to the tip of
              <varname>ref</varname>.
            </para>
          </listitem>
        </varlistentry>
        <varlistentry>
          <term>ref</term>
          <listitem>
            <para>
              The git ref to look for the requested revision under.
              This is often a branch or tag name. Defaults to
              <literal>HEAD</literal>.
            </para>

            <para>
              By default, the <varname>ref</varname> value is prefixed
              with <literal>refs/heads/</literal>. As of Nix 2.3.0
              Nix will not prefix <literal>refs/heads/</literal> if
              <varname>ref</varname> starts with <literal>refs/</literal>.
            </para>
          </listitem>
        </varlistentry>
        <varlistentry>
          <term>submodules</term>
          <listitem>
            <para>
              A Boolean parameter that specifies whether submodules
              should be checked out. Defaults to
              <literal>false</literal>.
            </para>
          </listitem>
        </varlistentry>
      </variablelist>

      <example>
        <title>Fetching a private repository over SSH</title>
        <programlisting>builtins.fetchGit {
  url = "git@github.com:my-secret/repository.git";
  ref = "master";
  rev = "adab8b916a45068c044658c4158d81878f9ed1c3";
}</programlisting>
      </example>

      <example>
        <title>Fetching an arbitrary ref</title>
        <programlisting>builtins.fetchGit {
  url = "https://github.com/NixOS/nix.git";
  ref = "refs/heads/0.5-release";
}</programlisting>
      </example>

      <example>
        <title>Fetching a repository's specific commit on an arbitrary branch</title>
        <para>
          If the revision you're looking for is in the default branch
          of the git repository you don't strictly need to specify
          the branch name in the <varname>ref</varname> attribute.
        </para>
        <para>
          However, if the revision you're looking for is in a future
          branch for the non-default branch you will need to specify
          the the <varname>ref</varname> attribute as well.
        </para>
        <programlisting>builtins.fetchGit {
  url = "https://github.com/nixos/nix.git";
  rev = "841fcbd04755c7a2865c51c1e2d3b045976b7452";
  ref = "1.11-maintenance";
}</programlisting>
        <note>
          <para>
            It is nice to always specify the branch which a revision
            belongs to. Without the branch being specified, the
            fetcher might fail if the default branch changes.
            Additionally, it can be confusing to try a commit from a
            non-default branch and see the fetch fail. If the branch
            is specified the fault is much more obvious.
          </para>
        </note>
      </example>

      <example>
        <title>Fetching a repository's specific commit on the default branch</title>
        <para>
          If the revision you're looking for is in the default branch
          of the git repository you may omit the
          <varname>ref</varname> attribute.
        </para>
        <programlisting>builtins.fetchGit {
  url = "https://github.com/nixos/nix.git";
  rev = "841fcbd04755c7a2865c51c1e2d3b045976b7452";
}</programlisting>
      </example>

      <example>
        <title>Fetching a tag</title>
        <programlisting>builtins.fetchGit {
  url = "https://github.com/nixos/nix.git";
  ref = "refs/tags/1.9";
}</programlisting>
      </example>

      <example>
        <title>Fetching the latest version of a remote branch</title>
        <para>
          <function>builtins.fetchGit</function> can behave impurely
           fetch the latest version of a remote branch.
        </para>
        <note><para>Nix will refetch the branch in accordance to
        <xref linkend="conf-tarball-ttl"/>.</para></note>
        <note><para>This behavior is disabled in
        <emphasis>Pure evaluation mode</emphasis>.</para></note>
        <programlisting>builtins.fetchGit {
  url = "ssh://git@github.com/nixos/nix.git";
  ref = "master";
}</programlisting>
      </example>
    </listitem>
  </varlistentry>


  <varlistentry><term><function>builtins.filter</function>
  <replaceable>f</replaceable> <replaceable>xs</replaceable></term>

    <listitem><para>Return a list consisting of the elements of
    <replaceable>xs</replaceable> for which the function
    <replaceable>f</replaceable> returns
    <literal>true</literal>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-filterSource">
    <term><function>builtins.filterSource</function>
    <replaceable>e1</replaceable> <replaceable>e2</replaceable></term>

    <listitem>

      <para>This function allows you to copy sources into the Nix
      store while filtering certain files.  For instance, suppose that
      you want to use the directory <filename>source-dir</filename> as
      an input to a Nix expression, e.g.

<programlisting>
stdenv.mkDerivation {
  ...
  src = ./source-dir;
}
</programlisting>

      However, if <filename>source-dir</filename> is a Subversion
      working copy, then all those annoying <filename>.svn</filename>
      subdirectories will also be copied to the store.  Worse, the
      contents of those directories may change a lot, causing lots of
      spurious rebuilds.  With <function>filterSource</function> you
      can filter out the <filename>.svn</filename> directories:

<programlisting>
  src = builtins.filterSource
    (path: type: type != "directory" || baseNameOf path != ".svn")
    ./source-dir;
</programlisting>

      </para>

      <para>Thus, the first argument <replaceable>e1</replaceable>
      must be a predicate function that is called for each regular
      file, directory or symlink in the source tree
      <replaceable>e2</replaceable>.  If the function returns
      <literal>true</literal>, the file is copied to the Nix store,
      otherwise it is omitted.  The function is called with two
      arguments.  The first is the full path of the file.  The second
      is a string that identifies the type of the file, which is
      either <literal>"regular"</literal>,
      <literal>"directory"</literal>, <literal>"symlink"</literal> or
      <literal>"unknown"</literal> (for other kinds of files such as
      device nodes or fifos &#x2014; but note that those cannot be copied to
      the Nix store, so if the predicate returns
      <literal>true</literal> for them, the copy will fail). If you
      exclude a directory, the entire corresponding subtree of
      <replaceable>e2</replaceable> will be excluded.</para>

    </listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-foldl-prime">
    <term><function>builtins.foldl&#x2019;</function>
    <replaceable>op</replaceable> <replaceable>nul</replaceable> <replaceable>list</replaceable></term>

    <listitem><para>Reduce a list by applying a binary operator, from
    left to right, e.g. <literal>foldl&#x2019; op nul [x0 x1 x2 ...] = op (op
    (op nul x0) x1) x2) ...</literal>. The operator is applied
    strictly, i.e., its arguments are evaluated first. For example,
    <literal>foldl&#x2019; (x: y: x + y) 0 [1 2 3]</literal> evaluates to
    6.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-functionArgs">
    <term><function>builtins.functionArgs</function>
    <replaceable>f</replaceable></term>

    <listitem><para>
    Return a set containing the names of the formal arguments expected
    by the function <replaceable>f</replaceable>.
    The value of each attribute is a Boolean denoting whether the corresponding
    argument has a default value.  For instance,
    <literal>functionArgs ({ x, y ? 123}: ...)  =  { x = false; y = true; }</literal>.
    </para>

    <para>"Formal argument" here refers to the attributes pattern-matched by
    the function.  Plain lambdas are not included, e.g.
    <literal>functionArgs (x: ...)  =  { }</literal>.
    </para></listitem>
  </varlistentry>


  <varlistentry xml:id="builtin-fromJSON">
    <term><function>builtins.fromJSON</function> <replaceable>e</replaceable></term>

    <listitem><para>Convert a JSON string to a Nix
    value. For example,

<programlisting>
builtins.fromJSON ''{"x": [1, 2, 3], "y": null}''
</programlisting>

    returns the value <literal>{ x = [ 1 2 3 ]; y = null;
    }</literal>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-genList">
    <term><function>builtins.genList</function>
    <replaceable>generator</replaceable> <replaceable>length</replaceable></term>

    <listitem><para>Generate list of size
    <replaceable>length</replaceable>, with each element
    <replaceable>i</replaceable> equal to the value returned by
    <replaceable>generator</replaceable> <literal>i</literal>. For
    example,

<programlisting>
builtins.genList (x: x * x) 5
</programlisting>

    returns the list <literal>[ 0 1 4 9 16 ]</literal>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-getAttr">
    <term><function>builtins.getAttr</function>
    <replaceable>s</replaceable> <replaceable>set</replaceable></term>

    <listitem><para><function>getAttr</function> returns the attribute
    named <replaceable>s</replaceable> from
    <replaceable>set</replaceable>.  Evaluation aborts if the
    attribute doesn&#x2019;t exist.  This is a dynamic version of the
    <literal>.</literal> operator, since <replaceable>s</replaceable>
    is an expression rather than an identifier.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-getEnv">
    <term><function>builtins.getEnv</function>
    <replaceable>s</replaceable></term>

    <listitem><para><function>getEnv</function> returns the value of
    the environment variable <replaceable>s</replaceable>, or an empty
    string if the variable doesn&#x2019;t exist.  This function should be
    used with care, as it can introduce all sorts of nasty environment
    dependencies in your Nix expression.</para>

    <para><function>getEnv</function> is used in Nix Packages to
    locate the file <filename>~/.nixpkgs/config.nix</filename>, which
    contains user-local settings for Nix Packages.  (That is, it does
    a <literal>getEnv "HOME"</literal> to locate the user&#x2019;s home
    directory.)</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-hasAttr">
    <term><function>builtins.hasAttr</function>
    <replaceable>s</replaceable> <replaceable>set</replaceable></term>

    <listitem><para><function>hasAttr</function> returns
    <literal>true</literal> if <replaceable>set</replaceable> has an
    attribute named <replaceable>s</replaceable>, and
    <literal>false</literal> otherwise.  This is a dynamic version of
    the <literal>?</literal>  operator, since
    <replaceable>s</replaceable> is an expression rather than an
    identifier.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-hashString">
    <term><function>builtins.hashString</function>
    <replaceable>type</replaceable> <replaceable>s</replaceable></term>

    <listitem><para>Return a base-16 representation of the
    cryptographic hash of string <replaceable>s</replaceable>.  The
    hash algorithm specified by <replaceable>type</replaceable> must
    be one of <literal>"md5"</literal>, <literal>"sha1"</literal>,
    <literal>"sha256"</literal> or <literal>"sha512"</literal>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-hashFile">
    <term><function>builtins.hashFile</function>
    <replaceable>type</replaceable> <replaceable>p</replaceable></term>

    <listitem><para>Return a base-16 representation of the
    cryptographic hash of the file at path <replaceable>p</replaceable>.  The
    hash algorithm specified by <replaceable>type</replaceable> must
    be one of <literal>"md5"</literal>, <literal>"sha1"</literal>,
    <literal>"sha256"</literal> or <literal>"sha512"</literal>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-head">
    <term><function>builtins.head</function>
    <replaceable>list</replaceable></term>

    <listitem><para>Return the first element of a list; abort
    evaluation if the argument isn&#x2019;t a list or is an empty list.  You
    can test whether a list is empty by comparing it with
    <literal>[]</literal>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-import">
    <term><function>import</function>
    <replaceable>path</replaceable></term>
    <term><function>builtins.import</function>
    <replaceable>path</replaceable></term>

    <listitem><para>Load, parse and return the Nix expression in the
    file <replaceable>path</replaceable>.  If <replaceable>path
    </replaceable> is a directory, the file <filename>default.nix
    </filename> in that directory is loaded.  Evaluation aborts if the
    file doesn&#x2019;t exist or contains an incorrect Nix expression.
    <function>import</function> implements Nix&#x2019;s module system: you
    can put any Nix expression (such as a set or a function) in a
    separate file, and use it from Nix expressions in other
    files.</para>

    <note><para>Unlike some languages, <function>import</function> is a regular
    function in Nix. Paths using the angle bracket syntax (e.g., <function>
    import</function> <replaceable>&lt;foo&gt;</replaceable>) are normal path
    values (see <xref linkend="ssec-values"/>).</para></note>

    <para>A Nix expression loaded by <function>import</function> must
    not contain any <emphasis>free variables</emphasis> (identifiers
    that are not defined in the Nix expression itself and are not
    built-in).  Therefore, it cannot refer to variables that are in
    scope at the call site.  For instance, if you have a calling
    expression

<programlisting>
rec {
  x = 123;
  y = import ./foo.nix;
}</programlisting>

    then the following <filename>foo.nix</filename> will give an
    error:

<programlisting>
x + 456</programlisting>

    since <varname>x</varname> is not in scope in
    <filename>foo.nix</filename>.  If you want <varname>x</varname>
    to be available in <filename>foo.nix</filename>, you should pass
    it as a function argument:

<programlisting>
rec {
  x = 123;
  y = import ./foo.nix x;
}</programlisting>

    and

<programlisting>
x: x + 456</programlisting>

    (The function argument doesn&#x2019;t have to be called
    <varname>x</varname> in <filename>foo.nix</filename>; any name
    would work.)</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-intersectAttrs">
    <term><function>builtins.intersectAttrs</function>
    <replaceable>e1</replaceable> <replaceable>e2</replaceable></term>

    <listitem><para>Return a set consisting of the attributes in the
    set <replaceable>e2</replaceable> that also exist in the set
    <replaceable>e1</replaceable>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-isAttrs">
    <term><function>builtins.isAttrs</function>
    <replaceable>e</replaceable></term>

    <listitem><para>Return <literal>true</literal> if
    <replaceable>e</replaceable> evaluates to a set, and
    <literal>false</literal> otherwise.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-isList">
    <term><function>builtins.isList</function>
    <replaceable>e</replaceable></term>

    <listitem><para>Return <literal>true</literal> if
    <replaceable>e</replaceable> evaluates to a list, and
    <literal>false</literal> otherwise.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-isFunction"><term><function>builtins.isFunction</function>
  <replaceable>e</replaceable></term>

    <listitem><para>Return <literal>true</literal> if
    <replaceable>e</replaceable> evaluates to a function, and
    <literal>false</literal> otherwise.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-isString">
    <term><function>builtins.isString</function>
    <replaceable>e</replaceable></term>

    <listitem><para>Return <literal>true</literal> if
    <replaceable>e</replaceable> evaluates to a string, and
    <literal>false</literal> otherwise.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-isInt">
    <term><function>builtins.isInt</function>
    <replaceable>e</replaceable></term>

    <listitem><para>Return <literal>true</literal> if
    <replaceable>e</replaceable> evaluates to an int, and
    <literal>false</literal> otherwise.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-isFloat">
    <term><function>builtins.isFloat</function>
    <replaceable>e</replaceable></term>

    <listitem><para>Return <literal>true</literal> if
    <replaceable>e</replaceable> evaluates to a float, and
    <literal>false</literal> otherwise.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-isBool">
    <term><function>builtins.isBool</function>
    <replaceable>e</replaceable></term>

    <listitem><para>Return <literal>true</literal> if
    <replaceable>e</replaceable> evaluates to a bool, and
    <literal>false</literal> otherwise.</para></listitem>

  </varlistentry>

  <varlistentry><term><function>builtins.isPath</function>
  <replaceable>e</replaceable></term>

    <listitem><para>Return <literal>true</literal> if
    <replaceable>e</replaceable> evaluates to a path, and
    <literal>false</literal> otherwise.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="builtin-isNull">
    <term><function>isNull</function>
    <replaceable>e</replaceable></term>
    <term><function>builtins.isNull</function>
    <replaceable>e</replaceable></term>

    <listitem><para>Return <literal>true</literal> if
    <replaceable>e</replaceable> evaluates to <literal>null</literal>,
    and <literal>false</literal> otherwise.</para>

    <warning><para>This function is <emphasis>deprecated</emphasis>;
    just write <literal>e == null</literal> instead.</para></warning>

    </listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-length">
    <term><function>builtins.length</function>
    <replaceable>e</replaceable></term>

    <listitem><para>Return the length of the list
    <replaceable>e</replaceable>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-lessThan">
    <term><function>builtins.lessThan</function>
    <replaceable>e1</replaceable> <replaceable>e2</replaceable></term>

    <listitem><para>Return <literal>true</literal> if the number
    <replaceable>e1</replaceable> is less than the number
    <replaceable>e2</replaceable>, and <literal>false</literal>
    otherwise.  Evaluation aborts if either
    <replaceable>e1</replaceable> or <replaceable>e2</replaceable>
    does not evaluate to a number.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-listToAttrs">
    <term><function>builtins.listToAttrs</function>
    <replaceable>e</replaceable></term>

    <listitem><para>Construct a set from a list specifying the names
    and values of each attribute.  Each element of the list should be
    a set consisting of a string-valued attribute
    <varname>name</varname> specifying the name of the attribute, and
    an attribute <varname>value</varname> specifying its value.
    Example:

<programlisting>
builtins.listToAttrs
  [ { name = "foo"; value = 123; }
    { name = "bar"; value = 456; }
  ]
</programlisting>

    evaluates to

<programlisting>
{ foo = 123; bar = 456; }
</programlisting>

    </para></listitem>

  </varlistentry>

  <varlistentry xml:id="builtin-map">
    <term><function>map</function>
    <replaceable>f</replaceable> <replaceable>list</replaceable></term>
    <term><function>builtins.map</function>
    <replaceable>f</replaceable> <replaceable>list</replaceable></term>

    <listitem><para>Apply the function <replaceable>f</replaceable> to
    each element in the list <replaceable>list</replaceable>.  For
    example,

<programlisting>
map (x: "foo" + x) [ "bar" "bla" "abc" ]</programlisting>

    evaluates to <literal>[ "foobar" "foobla" "fooabc"
    ]</literal>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-match">
    <term><function>builtins.match</function>
    <replaceable>regex</replaceable> <replaceable>str</replaceable></term>

    <listitem><para>Returns a list if the <link xlink:href="http://pubs.opengroup.org/onlinepubs/9699919799/basedefs/V1_chap09.html#tag_09_04">extended
    POSIX regular expression</link> <replaceable>regex</replaceable>
    matches <replaceable>str</replaceable> precisely, otherwise returns
    <literal>null</literal>.  Each item in the list is a regex group.

<programlisting>
builtins.match "ab" "abc"
</programlisting>

Evaluates to <literal>null</literal>.

<programlisting>
builtins.match "abc" "abc"
</programlisting>

Evaluates to <literal>[ ]</literal>.

<programlisting>
builtins.match "a(b)(c)" "abc"
</programlisting>

Evaluates to <literal>[ "b" "c" ]</literal>.

<programlisting>
builtins.match "[[:space:]]+([[:upper:]]+)[[:space:]]+" "  FOO   "
</programlisting>

Evaluates to <literal>[ "foo" ]</literal>.

    </para></listitem>
  </varlistentry>

  <varlistentry xml:id="builtin-mul">
    <term><function>builtins.mul</function>
    <replaceable>e1</replaceable> <replaceable>e2</replaceable></term>

    <listitem><para>Return the product of the numbers
    <replaceable>e1</replaceable> and
    <replaceable>e2</replaceable>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-parseDrvName">
    <term><function>builtins.parseDrvName</function>
    <replaceable>s</replaceable></term>

    <listitem><para>Split the string <replaceable>s</replaceable> into
    a package name and version.  The package name is everything up to
    but not including the first dash followed by a digit, and the
    version is everything following that dash.  The result is returned
    in a set <literal>{ name, version }</literal>.  Thus,
    <literal>builtins.parseDrvName "nix-0.12pre12876"</literal>
    returns <literal>{ name = "nix"; version = "0.12pre12876";
    }</literal>.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="builtin-path">
    <term>
      <function>builtins.path</function>
      <replaceable>args</replaceable>
    </term>

    <listitem>
      <para>
        An enrichment of the built-in path type, based on the attributes
        present in <replaceable>args</replaceable>. All are optional
        except <varname>path</varname>:
      </para>

      <variablelist>
        <varlistentry>
          <term>path</term>
          <listitem>
            <para>The underlying path.</para>
          </listitem>
        </varlistentry>
        <varlistentry>
          <term>name</term>
          <listitem>
            <para>
              The name of the path when added to the store. This can
              used to reference paths that have nix-illegal characters
              in their names, like <literal>@</literal>.
            </para>
          </listitem>
        </varlistentry>
        <varlistentry>
          <term>filter</term>
          <listitem>
            <para>
              A function of the type expected by
              <link linkend="builtin-filterSource">builtins.filterSource</link>,
              with the same semantics.
            </para>
          </listitem>
        </varlistentry>
        <varlistentry>
          <term>recursive</term>
          <listitem>
            <para>
              When <literal>false</literal>, when
              <varname>path</varname> is added to the store it is with a
              flat hash, rather than a hash of the NAR serialization of
              the file. Thus, <varname>path</varname> must refer to a
              regular file, not a directory. This allows similar
              behavior to <literal>fetchurl</literal>. Defaults to
              <literal>true</literal>.
            </para>
          </listitem>
        </varlistentry>
        <varlistentry>
          <term>sha256</term>
          <listitem>
            <para>
              When provided, this is the expected hash of the file at
              the path. Evaluation will fail if the hash is incorrect,
              and providing a hash allows
              <literal>builtins.path</literal> to be used even when the
              <literal>pure-eval</literal> nix config option is on.
            </para>
          </listitem>
        </varlistentry>
      </variablelist>
    </listitem>
  </varlistentry>

  <varlistentry xml:id="builtin-pathExists">
    <term><function>builtins.pathExists</function>
    <replaceable>path</replaceable></term>

    <listitem><para>Return <literal>true</literal> if the path
    <replaceable>path</replaceable> exists at evaluation time, and
    <literal>false</literal> otherwise.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="builtin-placeholder">
    <term><function>builtins.placeholder</function>
    <replaceable>output</replaceable></term>

    <listitem><para>Return a placeholder string for the specified
    <replaceable>output</replaceable> that will be substituted by the
    corresponding output path at build time. Typical outputs would be
    <literal>"out"</literal>, <literal>"bin"</literal> or
    <literal>"dev"</literal>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="builtin-readDir">
    <term><function>builtins.readDir</function>
    <replaceable>path</replaceable></term>

    <listitem><para>Return the contents of the directory
    <replaceable>path</replaceable> as a set mapping directory entries
    to the corresponding file type. For instance, if directory
    <filename>A</filename> contains a regular file
    <filename>B</filename> and another directory
    <filename>C</filename>, then <literal>builtins.readDir
    ./A</literal> will return the set

<programlisting>
{ B = "regular"; C = "directory"; }</programlisting>

    The possible values for the file type are
    <literal>"regular"</literal>, <literal>"directory"</literal>,
    <literal>"symlink"</literal> and
    <literal>"unknown"</literal>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-readFile">
    <term><function>builtins.readFile</function>
    <replaceable>path</replaceable></term>

    <listitem><para>Return the contents of the file
    <replaceable>path</replaceable> as a string.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-removeAttrs">
    <term><function>removeAttrs</function>
    <replaceable>set</replaceable> <replaceable>list</replaceable></term>
    <term><function>builtins.removeAttrs</function>
    <replaceable>set</replaceable> <replaceable>list</replaceable></term>

    <listitem><para>Remove the attributes listed in
    <replaceable>list</replaceable> from
    <replaceable>set</replaceable>.  The attributes don&#x2019;t have to
    exist in <replaceable>set</replaceable>. For instance,

<programlisting>
removeAttrs { x = 1; y = 2; z = 3; } [ "a" "x" "z" ]</programlisting>

    evaluates to <literal>{ y = 2; }</literal>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-replaceStrings">
    <term><function>builtins.replaceStrings</function>
    <replaceable>from</replaceable> <replaceable>to</replaceable> <replaceable>s</replaceable></term>

    <listitem><para>Given string <replaceable>s</replaceable>, replace
    every occurrence of the strings in <replaceable>from</replaceable>
    with the corresponding string in
    <replaceable>to</replaceable>. For example,

<programlisting>
builtins.replaceStrings ["oo" "a"] ["a" "i"] "foobar"
</programlisting>

    evaluates to <literal>"fabir"</literal>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-seq">
    <term><function>builtins.seq</function>
    <replaceable>e1</replaceable> <replaceable>e2</replaceable></term>

    <listitem><para>Evaluate <replaceable>e1</replaceable>, then
    evaluate and return <replaceable>e2</replaceable>. This ensures
    that a computation is strict in the value of
    <replaceable>e1</replaceable>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-sort">
    <term><function>builtins.sort</function>
    <replaceable>comparator</replaceable> <replaceable>list</replaceable></term>

    <listitem><para>Return <replaceable>list</replaceable> in sorted
    order. It repeatedly calls the function
    <replaceable>comparator</replaceable> with two elements. The
    comparator should return <literal>true</literal> if the first
    element is less than the second, and <literal>false</literal>
    otherwise. For example,

<programlisting>
builtins.sort builtins.lessThan [ 483 249 526 147 42 77 ]
</programlisting>

    produces the list <literal>[ 42 77 147 249 483 526
    ]</literal>.</para>

    <para>This is a stable sort: it preserves the relative order of
    elements deemed equal by the comparator.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-split">
    <term><function>builtins.split</function>
    <replaceable>regex</replaceable> <replaceable>str</replaceable></term>

    <listitem><para>Returns a list composed of non matched strings interleaved
    with the lists of the <link xlink:href="http://pubs.opengroup.org/onlinepubs/9699919799/basedefs/V1_chap09.html#tag_09_04">extended
    POSIX regular expression</link> <replaceable>regex</replaceable> matches
    of <replaceable>str</replaceable>. Each item in the lists of matched
    sequences is a regex group.

<programlisting>
builtins.split "(a)b" "abc"
</programlisting>

Evaluates to <literal>[ "" [ "a" ] "c" ]</literal>.

<programlisting>
builtins.split "([ac])" "abc"
</programlisting>

Evaluates to <literal>[ "" [ "a" ] "b" [ "c" ] "" ]</literal>.

<programlisting>
builtins.split "(a)|(c)" "abc"
</programlisting>

Evaluates to <literal>[ "" [ "a" null ] "b" [ null "c" ] "" ]</literal>.

<programlisting>
builtins.split "([[:upper:]]+)" "  FOO   "
</programlisting>

Evaluates to <literal>[ "  " [ "FOO" ] "   " ]</literal>.

    </para></listitem>
  </varlistentry>


  <varlistentry xml:id="builtin-splitVersion">
    <term><function>builtins.splitVersion</function>
    <replaceable>s</replaceable></term>

    <listitem><para>Split a string representing a version into its
    components, by the same version splitting logic underlying the
    version comparison in <link linkend="ssec-version-comparisons">
    <command>nix-env -u</command></link>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-stringLength">
    <term><function>builtins.stringLength</function>
    <replaceable>e</replaceable></term>

    <listitem><para>Return the length of the string
    <replaceable>e</replaceable>.  If <replaceable>e</replaceable> is
    not a string, evaluation is aborted.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-sub">
    <term><function>builtins.sub</function>
    <replaceable>e1</replaceable> <replaceable>e2</replaceable></term>

    <listitem><para>Return the difference between the numbers
    <replaceable>e1</replaceable> and
    <replaceable>e2</replaceable>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-substring">
    <term><function>builtins.substring</function>
    <replaceable>start</replaceable> <replaceable>len</replaceable>
    <replaceable>s</replaceable></term>

    <listitem><para>Return the substring of
    <replaceable>s</replaceable> from character position
    <replaceable>start</replaceable> (zero-based) up to but not
    including <replaceable>start + len</replaceable>.  If
    <replaceable>start</replaceable> is greater than the length of the
    string, an empty string is returned, and if <replaceable>start +
    len</replaceable> lies beyond the end of the string, only the
    substring up to the end of the string is returned.
    <replaceable>start</replaceable> must be
    non-negative. For example,

<programlisting>
builtins.substring 0 3 "nixos"
</programlisting>

   evaluates to <literal>"nix"</literal>.
   </para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-tail">
    <term><function>builtins.tail</function>
    <replaceable>list</replaceable></term>

    <listitem><para>Return the second to last elements of a list;
    abort evaluation if the argument isn&#x2019;t a list or is an empty
    list.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-throw">
    <term><function>throw</function>
    <replaceable>s</replaceable></term>
    <term><function>builtins.throw</function>
    <replaceable>s</replaceable></term>

    <listitem><para>Throw an error message
    <replaceable>s</replaceable>.  This usually aborts Nix expression
    evaluation, but in <command>nix-env -qa</command> and other
    commands that try to evaluate a set of derivations to get
    information about those derivations, a derivation that throws an
    error is silently skipped (which is not the case for
    <function>abort</function>).</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-toFile">
    <term><function>builtins.toFile</function>
    <replaceable>name</replaceable>
    <replaceable>s</replaceable></term>

    <listitem><para>Store the string <replaceable>s</replaceable> in a
    file in the Nix store and return its path.  The file has suffix
    <replaceable>name</replaceable>.  This file can be used as an
    input to derivations.  One application is to write builders
    &#x201C;inline&#x201D;.  For instance, the following Nix expression combines
    <xref linkend="ex-hello-nix"/> and <xref linkend="ex-hello-builder"/> into one file:

<programlisting>
{ stdenv, fetchurl, perl }:

stdenv.mkDerivation {
  name = "hello-2.1.1";

  builder = builtins.toFile "builder.sh" "
    source $stdenv/setup

    PATH=$perl/bin:$PATH

    tar xvfz $src
    cd hello-*
    ./configure --prefix=$out
    make
    make install
  ";

  src = fetchurl {
    url = "http://ftp.nluug.nl/pub/gnu/hello/hello-2.1.1.tar.gz";
    sha256 = "1md7jsfd8pa45z73bz1kszpp01yw6x5ljkjk2hx7wl800any6465";
  };
  inherit perl;
}</programlisting>

    </para>

    <para>It is even possible for one file to refer to another, e.g.,

<programlisting>
  builder = let
    configFile = builtins.toFile "foo.conf" "
      # This is some dummy configuration file.
      <replaceable>...</replaceable>
    ";
  in builtins.toFile "builder.sh" "
    source $stdenv/setup
    <replaceable>...</replaceable>
    cp ${configFile} $out/etc/foo.conf
  ";</programlisting>

    Note that <literal>${configFile}</literal> is an antiquotation
    (see <xref linkend="ssec-values"/>), so the result of the
    expression <literal>configFile</literal> (i.e., a path like
    <filename>/nix/store/m7p7jfny445k...-foo.conf</filename>) will be
    spliced into the resulting string.</para>

    <para>It is however <emphasis>not</emphasis> allowed to have files
    mutually referring to each other, like so:

<programlisting>
let
  foo = builtins.toFile "foo" "...${bar}...";
  bar = builtins.toFile "bar" "...${foo}...";
in foo</programlisting>

    This is not allowed because it would cause a cyclic dependency in
    the computation of the cryptographic hashes for
    <varname>foo</varname> and <varname>bar</varname>.</para>
    <para>It is also not possible to reference the result of a derivation.
    If you are using Nixpkgs, the <literal>writeTextFile</literal> function is able to
    do that.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-toJSON">
    <term><function>builtins.toJSON</function> <replaceable>e</replaceable></term>

    <listitem><para>Return a string containing a JSON representation
    of <replaceable>e</replaceable>.  Strings, integers, floats, booleans,
    nulls and lists are mapped to their JSON equivalents.  Sets
    (except derivations) are represented as objects.  Derivations are
    translated to a JSON string containing the derivation&#x2019;s output
    path.  Paths are copied to the store and represented as a JSON
    string of the resulting store path.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-toPath">
    <term><function>builtins.toPath</function> <replaceable>s</replaceable></term>

    <listitem><para> DEPRECATED. Use <literal>/. + "/path"</literal>
    to convert a string into an absolute path. For relative paths,
    use <literal>./. + "/path"</literal>.
    </para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-toString">
    <term><function>toString</function> <replaceable>e</replaceable></term>
    <term><function>builtins.toString</function> <replaceable>e</replaceable></term>

    <listitem><para>Convert the expression
    <replaceable>e</replaceable> to a string.
    <replaceable>e</replaceable> can be:</para>
    <itemizedlist>
      <listitem><para>A string (in which case the string is returned unmodified).</para></listitem>
      <listitem><para>A path (e.g., <literal>toString /foo/bar</literal> yields <literal>"/foo/bar"</literal>.</para></listitem>
      <listitem><para>A set containing <literal>{ __toString = self: ...; }</literal>.</para></listitem>
      <listitem><para>An integer.</para></listitem>
      <listitem><para>A list, in which case the string representations of its elements are joined with spaces.</para></listitem>
      <listitem><para>A Boolean (<literal>false</literal> yields <literal>""</literal>, <literal>true</literal> yields <literal>"1"</literal>).</para></listitem>
      <listitem><para><literal>null</literal>, which yields the empty string.</para></listitem>
    </itemizedlist>
    </listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-toXML">
    <term><function>builtins.toXML</function> <replaceable>e</replaceable></term>

    <listitem><para>Return a string containing an XML representation
    of <replaceable>e</replaceable>.  The main application for
    <function>toXML</function> is to communicate information with the
    builder in a more structured format than plain environment
    variables.</para>

    <!-- TODO: more formally describe the schema of the XML
    representation -->

    <para><xref linkend="ex-toxml"/> shows an example where this is
    the case.  The builder is supposed to generate the configuration
    file for a <link xlink:href="http://jetty.mortbay.org/">Jetty
    servlet container</link>.  A servlet container contains a number
    of servlets (<filename>*.war</filename> files) each exported under
    a specific URI prefix.  So the servlet configuration is a list of
    sets containing the <varname>path</varname> and
    <varname>war</varname> of the servlet (<xref linkend="ex-toxml-co-servlets"/>).  This kind of information is
    difficult to communicate with the normal method of passing
    information through an environment variable, which just
    concatenates everything together into a string (which might just
    work in this case, but wouldn&#x2019;t work if fields are optional or
    contain lists themselves).  Instead the Nix expression is
    converted to an XML representation with
    <function>toXML</function>, which is unambiguous and can easily be
    processed with the appropriate tools.  For instance, in the
    example an XSLT stylesheet (<xref linkend="ex-toxml-co-stylesheet"/>) is applied to it (<xref linkend="ex-toxml-co-apply"/>) to
    generate the XML configuration file for the Jetty server.  The XML
    representation produced from <xref linkend="ex-toxml-co-servlets"/> by <function>toXML</function> is shown in <xref linkend="ex-toxml-result"/>.</para>

    <para>Note that <xref linkend="ex-toxml"/> uses the <function linkend="builtin-toFile">toFile</function> built-in to write the
    builder and the stylesheet &#x201C;inline&#x201D; in the Nix expression.  The
    path of the stylesheet is spliced into the builder at
    <literal>xsltproc ${stylesheet}
    <replaceable>...</replaceable></literal>.</para>

    <example xml:id="ex-toxml"><title>Passing information to a builder
    using <function>toXML</function></title>

<programlisting><![CDATA[
{ stdenv, fetchurl, libxslt, jira, uberwiki }:

stdenv.mkDerivation (rec {
  name = "web-server";

  buildInputs = [ libxslt ];

  builder = builtins.toFile "builder.sh" "
    source $stdenv/setup
    mkdir $out
    echo "$servlets" | xsltproc ${stylesheet} - > $out/server-conf.xml]]> <co xml:id="ex-toxml-co-apply"/> <![CDATA[
  ";

  stylesheet = builtins.toFile "stylesheet.xsl"]]> <co xml:id="ex-toxml-co-stylesheet"/> <![CDATA[
   "<?xml version='1.0' encoding='UTF-8'?>
    <xsl:stylesheet xmlns:xsl='http://www.w3.org/1999/XSL/Transform' version='1.0'>
      <xsl:template match='/'>
        <Configure>
          <xsl:for-each select='/expr/list/attrs'>
            <Call name='addWebApplication'>
              <Arg><xsl:value-of select=\"attr[@name = 'path']/string/@value\" /></Arg>
              <Arg><xsl:value-of select=\"attr[@name = 'war']/path/@value\" /></Arg>
            </Call>
          </xsl:for-each>
        </Configure>
      </xsl:template>
    </xsl:stylesheet>
  ";

  servlets = builtins.toXML []]> <co xml:id="ex-toxml-co-servlets"/> <![CDATA[
    { path = "/bugtracker"; war = jira + "/lib/atlassian-jira.war"; }
    { path = "/wiki"; war = uberwiki + "/uberwiki.war"; }
  ];
})]]></programlisting>

    </example>

    <example xml:id="ex-toxml-result"><title>XML representation produced by
    <function>toXML</function></title>

<programlisting><![CDATA[<?xml version='1.0' encoding='utf-8'?>
<expr>
  <list>
    <attrs>
      <attr name="path">
        <string value="/bugtracker" />
      </attr>
      <attr name="war">
        <path value="/nix/store/d1jh9pasa7k2...-jira/lib/atlassian-jira.war" />
      </attr>
    </attrs>
    <attrs>
      <attr name="path">
        <string value="/wiki" />
      </attr>
      <attr name="war">
        <path value="/nix/store/y6423b1yi4sx...-uberwiki/uberwiki.war" />
      </attr>
    </attrs>
  </list>
</expr>]]></programlisting>

    </example>

    </listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-trace">
    <term><function>builtins.trace</function>
    <replaceable>e1</replaceable> <replaceable>e2</replaceable></term>

    <listitem><para>Evaluate <replaceable>e1</replaceable> and print its
    abstract syntax representation on standard error.  Then return
    <replaceable>e2</replaceable>.  This function is useful for
    debugging.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="builtin-tryEval">
    <term><function>builtins.tryEval</function>
    <replaceable>e</replaceable></term>

    <listitem><para>Try to shallowly evaluate <replaceable>e</replaceable>.
    Return a set containing the attributes <literal>success</literal>
    (<literal>true</literal> if <replaceable>e</replaceable> evaluated
    successfully, <literal>false</literal> if an error was thrown) and
    <literal>value</literal>, equalling <replaceable>e</replaceable>
    if successful and <literal>false</literal> otherwise. Note that this
    doesn't evaluate <replaceable>e</replaceable> deeply, so
    <literal>let e = { x = throw ""; }; in (builtins.tryEval e).success
    </literal> will be <literal>true</literal>. Using <literal>builtins.deepSeq
    </literal> one can get the expected result: <literal>let e = { x = throw "";
    }; in (builtins.tryEval (builtins.deepSeq e e)).success</literal> will be
    <literal>false</literal>.
    </para></listitem>

  </varlistentry>


  <varlistentry xml:id="builtin-typeOf">
    <term><function>builtins.typeOf</function>
    <replaceable>e</replaceable></term>

    <listitem><para>Return a string representing the type of the value
    <replaceable>e</replaceable>, namely <literal>"int"</literal>,
    <literal>"bool"</literal>, <literal>"string"</literal>,
    <literal>"path"</literal>, <literal>"null"</literal>,
    <literal>"set"</literal>, <literal>"list"</literal>,
    <literal>"lambda"</literal> or
    <literal>"float"</literal>.</para></listitem>

  </varlistentry>


</variablelist>


</section>


</chapter>

</part>
  <part xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" xml:id="part-advanced-topics" version="5.0" xml:base="advanced-topics/advanced-topics.xml">

<title>Advanced Topics</title>

<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="chap-distributed-builds">

<title>Remote Builds</title>

<para>Nix supports remote builds, where a local Nix installation can
forward Nix builds to other machines.  This allows multiple builds to
be performed in parallel and allows Nix to perform multi-platform
builds in a semi-transparent way.  For instance, if you perform a
build for a <literal>x86_64-darwin</literal> on an
<literal>i686-linux</literal> machine, Nix can automatically forward
the build to a <literal>x86_64-darwin</literal> machine, if
available.</para>

<para>To forward a build to a remote machine, it&#x2019;s required that the
remote machine is accessible via SSH and that it has Nix
installed. You can test whether connecting to the remote Nix instance
works, e.g.

<screen>
$ nix ping-store --store ssh://mac
</screen>

will try to connect to the machine named <literal>mac</literal>. It is
possible to specify an SSH identity file as part of the remote store
URI, e.g.

<screen>
$ nix ping-store --store ssh://mac?ssh-key=/home/alice/my-key
</screen>

Since builds should be non-interactive, the key should not have a
passphrase. Alternatively, you can load identities ahead of time into
<command>ssh-agent</command> or <command>gpg-agent</command>.</para>

<para>If you get the error

<screen>
bash: nix-store: command not found
error: cannot connect to 'mac'
</screen>

then you need to ensure that the <envar>PATH</envar> of
non-interactive login shells contains Nix.</para>

<warning><para>If you are building via the Nix daemon, it is the Nix
daemon user account (that is, <literal>root</literal>) that should
have SSH access to the remote machine. If you can&#x2019;t or don&#x2019;t want to
configure <literal>root</literal> to be able to access to remote
machine, you can use a private Nix store instead by passing
e.g. <literal>--store ~/my-nix</literal>.</para></warning>

<para>The list of remote machines can be specified on the command line
or in the Nix configuration file. The former is convenient for
testing. For example, the following command allows you to build a
derivation for <literal>x86_64-darwin</literal> on a Linux machine:

<screen>
$ uname
Linux

$ nix build \
  '(with import &lt;nixpkgs&gt; { system = "x86_64-darwin"; }; runCommand "foo" {} "uname &gt; $out")' \
  --builders 'ssh://mac x86_64-darwin'
[1/0/1 built, 0.0 MiB DL] building foo on ssh://mac

$ cat ./result
Darwin
</screen>

It is possible to specify multiple builders separated by a semicolon
or a newline, e.g.

<screen>
  --builders 'ssh://mac x86_64-darwin ; ssh://beastie x86_64-freebsd'
</screen>
</para>

<para>Each machine specification consists of the following elements,
separated by spaces. Only the first element is required.
To leave a field at its default, set it to <literal>-</literal>.

<orderedlist>

  <listitem><para>The URI of the remote store in the format
  <literal>ssh://[<replaceable>username</replaceable>@]<replaceable>hostname</replaceable></literal>,
  e.g. <literal>ssh://nix@mac</literal> or
  <literal>ssh://mac</literal>. For backward compatibility,
  <literal>ssh://</literal> may be omitted. The hostname may be an
  alias defined in your
  <filename>~/.ssh/config</filename>.</para></listitem>

  <listitem><para>A comma-separated list of Nix platform type
  identifiers, such as <literal>x86_64-darwin</literal>.  It is
  possible for a machine to support multiple platform types, e.g.,
  <literal>i686-linux,x86_64-linux</literal>. If omitted, this
  defaults to the local platform type.</para></listitem>

  <listitem><para>The SSH identity file to be used to log in to the
  remote machine. If omitted, SSH will use its regular
  identities.</para></listitem>

  <listitem><para>The maximum number of builds that Nix will execute
  in parallel on the machine.  Typically this should be equal to the
  number of CPU cores.  For instance, the machine
  <literal>itchy</literal> in the example will execute up to 8 builds
  in parallel.</para></listitem>

  <listitem><para>The &#x201C;speed factor&#x201D;, indicating the relative speed of
  the machine.  If there are multiple machines of the right type, Nix
  will prefer the fastest, taking load into account.</para></listitem>

  <listitem><para>A comma-separated list of <emphasis>supported
  features</emphasis>.  If a derivation has the
  <varname>requiredSystemFeatures</varname> attribute, then Nix will
  only perform the derivation on a machine that has the specified
  features.  For instance, the attribute

<programlisting>
requiredSystemFeatures = [ "kvm" ];
</programlisting>

  will cause the build to be performed on a machine that has the
  <literal>kvm</literal> feature.</para></listitem>

  <listitem><para>A comma-separated list of <emphasis>mandatory
  features</emphasis>.  A machine will only be used to build a
  derivation if all of the machine&#x2019;s mandatory features appear in the
  derivation&#x2019;s <varname>requiredSystemFeatures</varname>
  attribute..</para></listitem>

</orderedlist>

For example, the machine specification

<programlisting>
nix@scratchy.labs.cs.uu.nl  i686-linux      /home/nix/.ssh/id_scratchy_auto        8 1 kvm
nix@itchy.labs.cs.uu.nl     i686-linux      /home/nix/.ssh/id_scratchy_auto        8 2
nix@poochie.labs.cs.uu.nl   i686-linux      /home/nix/.ssh/id_scratchy_auto        1 2 kvm benchmark
</programlisting>

specifies several machines that can perform
<literal>i686-linux</literal> builds. However,
<literal>poochie</literal> will only do builds that have the attribute

<programlisting>
requiredSystemFeatures = [ "benchmark" ];
</programlisting>

or

<programlisting>
requiredSystemFeatures = [ "benchmark" "kvm" ];
</programlisting>

<literal>itchy</literal> cannot do builds that require
<literal>kvm</literal>, but <literal>scratchy</literal> does support
such builds. For regular builds, <literal>itchy</literal> will be
preferred over <literal>scratchy</literal> because it has a higher
speed factor.</para>

<para>Remote builders can also be configured in
<filename>nix.conf</filename>, e.g.

<programlisting>
builders = ssh://mac x86_64-darwin ; ssh://beastie x86_64-freebsd
</programlisting>

Finally, remote builders can be configured in a separate configuration
file included in <option>builders</option> via the syntax
<literal>@<replaceable>file</replaceable></literal>. For example,

<programlisting>
builders = @/etc/nix/machines
</programlisting>

causes the list of machines in <filename>/etc/nix/machines</filename>
to be included. (This is the default.)</para>

<para>If you want the builders to use caches, you likely want to set
the option <link linkend="conf-builders-use-substitutes"><literal>builders-use-substitutes</literal></link>
in your local <filename>nix.conf</filename>.</para>

<para>To build only on remote builders and disable building on the local machine,
you can use the option <option>--max-jobs 0</option>.</para>

</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="chap-tuning-cores-and-jobs">

<title>Tuning Cores and Jobs</title>

<para>Nix has two relevant settings with regards to how your CPU cores
will be utilized: <xref linkend="conf-cores"/> and
<xref linkend="conf-max-jobs"/>. This chapter will talk about what
they are, how they interact, and their configuration trade-offs.</para>

<variablelist>
  <varlistentry>
    <term><xref linkend="conf-max-jobs"/></term>
    <listitem><para>
      Dictates how many separate derivations will be built at the same
      time. If you set this to zero, the local machine will do no
      builds. Nix will still substitute from binary caches, and build
      remotely if remote builders are configured.
    </para></listitem>
  </varlistentry>
  <varlistentry>
    <term><xref linkend="conf-cores"/></term>
    <listitem><para>
      Suggests how many cores each derivation should use. Similar to
      <command>make -j</command>.
    </para></listitem>
  </varlistentry>
</variablelist>

<para>The <xref linkend="conf-cores"/> setting determines the value of
<envar>NIX_BUILD_CORES</envar>. <envar>NIX_BUILD_CORES</envar> is equal
to <xref linkend="conf-cores"/>, unless <xref linkend="conf-cores"/>
equals <literal>0</literal>, in which case <envar>NIX_BUILD_CORES</envar>
will be the total number of cores in the system.</para>

<para>The maximum number of consumed cores is a simple multiplication,
<xref linkend="conf-max-jobs"/> * <envar>NIX_BUILD_CORES</envar>.</para>

<para>The balance on how to set these two independent variables depends
upon each builder's workload and hardware. Here are a few example
scenarios on a machine with 24 cores:</para>

<table>
  <caption>Balancing 24 Build Cores</caption>
  <thead>
    <tr>
      <th><xref linkend="conf-max-jobs"/></th>
      <th><xref linkend="conf-cores"/></th>
      <th><envar>NIX_BUILD_CORES</envar></th>
      <th>Maximum Processes</th>
      <th>Result</th>
    </tr>
  </thead>
  <tbody>
    <tr>
      <td>1</td>
      <td>24</td>
      <td>24</td>
      <td>24</td>
      <td>
        One derivation will be built at a time, each one can use 24
        cores. Undersold if a job can&#x2019;t use 24 cores.
      </td>
    </tr>

    <tr>
      <td>4</td>
      <td>6</td>
      <td>6</td>
      <td>24</td>
      <td>
        Four derivations will be built at once, each given access to
        six cores.
      </td>
    </tr>
    <tr>
      <td>12</td>
      <td>6</td>
      <td>6</td>
      <td>72</td>
      <td>
        12 derivations will be built at once, each given access to six
        cores. This configuration is over-sold. If all 12 derivations
        being built simultaneously try to use all six cores, the
        machine's performance will be degraded due to extensive context
        switching between the 12 builds.
      </td>
    </tr>
    <tr>
      <td>24</td>
      <td>1</td>
      <td>1</td>
      <td>24</td>
      <td>
        24 derivations can build at the same time, each using a single
        core. Never oversold, but derivations which require many cores
        will be very slow to compile.
      </td>
    </tr>
    <tr>
      <td>24</td>
      <td>0</td>
      <td>24</td>
      <td>576</td>
      <td>
        24 derivations can build at the same time, each using all the
        available cores of the machine. Very likely to be oversold,
        and very likely to suffer context switches.
      </td>
    </tr>
  </tbody>
</table>

<para>It is up to the derivations' build script to respect
host's requested cores-per-build by following the value of the
<envar>NIX_BUILD_CORES</envar> environment variable.</para>

</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" xml:id="chap-diff-hook" version="5.0">

<title>Verifying Build Reproducibility with <option linkend="conf-diff-hook">diff-hook</option></title>

<subtitle>Check build reproducibility by running builds multiple times
and comparing their results.</subtitle>

<para>Specify a program with Nix's <xref linkend="conf-diff-hook"/> to
compare build results when two builds produce different results. Note:
this hook is only executed if the results are not the same, this hook
is not used for determining if the results are the same.</para>

<para>For purposes of demonstration, we'll use the following Nix file,
<filename>deterministic.nix</filename> for testing:</para>

<programlisting>
let
  inherit (import &lt;nixpkgs&gt; {}) runCommand;
in {
  stable = runCommand "stable" {} ''
    touch $out
  '';

  unstable = runCommand "unstable" {} ''
    echo $RANDOM &gt; $out
  '';
}
</programlisting>

<para>Additionally, <filename>nix.conf</filename> contains:

<programlisting>
diff-hook = /etc/nix/my-diff-hook
run-diff-hook = true
</programlisting>

where <filename>/etc/nix/my-diff-hook</filename> is an executable
file containing:

<programlisting>
#!/bin/sh
exec &gt;&amp;2
echo "For derivation $3:"
/run/current-system/sw/bin/diff -r "$1" "$2"
</programlisting>

</para>

<para>The diff hook is executed by the same user and group who ran the
build. However, the diff hook does not have write access to the store
path just built.</para>

<section>
  <title>
    Spot-Checking Build Determinism
  </title>

  <para>
    Verify a path which already exists in the Nix store by passing
    <option>--check</option> to the build command.
  </para>

  <para>If the build passes and is deterministic, Nix will exit with a
  status code of 0:</para>

  <screen>
$ nix-build ./deterministic.nix -A stable
this derivation will be built:
  /nix/store/z98fasz2jqy9gs0xbvdj939p27jwda38-stable.drv
building '/nix/store/z98fasz2jqy9gs0xbvdj939p27jwda38-stable.drv'...
/nix/store/yyxlzw3vqaas7wfp04g0b1xg51f2czgq-stable

$ nix-build ./deterministic.nix -A stable --check
checking outputs of '/nix/store/z98fasz2jqy9gs0xbvdj939p27jwda38-stable.drv'...
/nix/store/yyxlzw3vqaas7wfp04g0b1xg51f2czgq-stable
</screen>

  <para>If the build is not deterministic, Nix will exit with a status
  code of 1:</para>

  <screen>
$ nix-build ./deterministic.nix -A unstable
this derivation will be built:
  /nix/store/cgl13lbj1w368r5z8gywipl1ifli7dhk-unstable.drv
building '/nix/store/cgl13lbj1w368r5z8gywipl1ifli7dhk-unstable.drv'...
/nix/store/krpqk0l9ib0ibi1d2w52z293zw455cap-unstable

$ nix-build ./deterministic.nix -A unstable --check
checking outputs of '/nix/store/cgl13lbj1w368r5z8gywipl1ifli7dhk-unstable.drv'...
error: derivation '/nix/store/cgl13lbj1w368r5z8gywipl1ifli7dhk-unstable.drv' may not be deterministic: output '/nix/store/krpqk0l9ib0ibi1d2w52z293zw455cap-unstable' differs
</screen>

<para>In the Nix daemon's log, we will now see:
<screen>
For derivation /nix/store/cgl13lbj1w368r5z8gywipl1ifli7dhk-unstable.drv:
1c1
&lt; 8108
---
&gt; 30204
</screen>
</para>

  <para>Using <option>--check</option> with <option>--keep-failed</option>
  will cause Nix to keep the second build's output in a special,
  <literal>.check</literal> path:</para>

  <screen>
$ nix-build ./deterministic.nix -A unstable --check --keep-failed
checking outputs of '/nix/store/cgl13lbj1w368r5z8gywipl1ifli7dhk-unstable.drv'...
note: keeping build directory '/tmp/nix-build-unstable.drv-0'
error: derivation '/nix/store/cgl13lbj1w368r5z8gywipl1ifli7dhk-unstable.drv' may not be deterministic: output '/nix/store/krpqk0l9ib0ibi1d2w52z293zw455cap-unstable' differs from '/nix/store/krpqk0l9ib0ibi1d2w52z293zw455cap-unstable.check'
</screen>

  <para>In particular, notice the
  <literal>/nix/store/krpqk0l9ib0ibi1d2w52z293zw455cap-unstable.check</literal>
  output. Nix has copied the build results to that directory where you
  can examine it.</para>

  <note xml:id="check-dirs-are-unregistered">
    <title><literal>.check</literal> paths are not registered store paths</title>

    <para>Check paths are not protected against garbage collection,
    and this path will be deleted on the next garbage collection.</para>

    <para>The path is guaranteed to be alive for the duration of
    <xref linkend="conf-diff-hook"/>'s execution, but may be deleted
    any time after.</para>

    <para>If the comparison is performed as part of automated tooling,
    please use the diff-hook or author your tooling to handle the case
    where the build was not deterministic and also a check path does
    not exist.</para>
  </note>

  <para>
    <option>--check</option> is only usable if the derivation has
    been built on the system already. If the derivation has not been
    built Nix will fail with the error:
    <screen>
error: some outputs of '/nix/store/hzi1h60z2qf0nb85iwnpvrai3j2w7rr6-unstable.drv' are not valid, so checking is not possible
</screen>

    Run the build without <option>--check</option>, and then try with
    <option>--check</option> again.
  </para>
</section>

<section>
  <title>
    Automatic and Optionally Enforced Determinism Verification
  </title>

  <para>
    Automatically verify every build at build time by executing the
    build multiple times.
  </para>

  <para>
    Setting <xref linkend="conf-repeat"/> and
    <xref linkend="conf-enforce-determinism"/> in your
    <filename>nix.conf</filename> permits the automated verification
    of every build Nix performs.
  </para>

  <para>
    The following configuration will run each build three times, and
    will require the build to be deterministic:

    <programlisting>
enforce-determinism = true
repeat = 2
</programlisting>
  </para>

  <para>
    Setting <xref linkend="conf-enforce-determinism"/> to false as in
    the following configuration will run the build multiple times,
    execute the build hook, but will allow the build to succeed even
    if it does not build reproducibly:

    <programlisting>
enforce-determinism = false
repeat = 1
</programlisting>
  </para>

  <para>
    An example output of this configuration:
    <screen>
$ nix-build ./test.nix -A unstable
this derivation will be built:
  /nix/store/ch6llwpr2h8c3jmnf3f2ghkhx59aa97f-unstable.drv
building '/nix/store/ch6llwpr2h8c3jmnf3f2ghkhx59aa97f-unstable.drv' (round 1/2)...
building '/nix/store/ch6llwpr2h8c3jmnf3f2ghkhx59aa97f-unstable.drv' (round 2/2)...
output '/nix/store/6xg356v9gl03hpbbg8gws77n19qanh02-unstable' of '/nix/store/ch6llwpr2h8c3jmnf3f2ghkhx59aa97f-unstable.drv' differs from '/nix/store/6xg356v9gl03hpbbg8gws77n19qanh02-unstable.check' from previous round
/nix/store/6xg356v9gl03hpbbg8gws77n19qanh02-unstable
</screen>
  </para>
</section>
</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" xml:id="chap-post-build-hook" version="5.0">

<title>Using the <option linkend="conf-post-build-hook">post-build-hook</option></title>
<subtitle>Uploading to an S3-compatible binary cache after each build</subtitle>


<section xml:id="chap-post-build-hook-caveats">
  <title>Implementation Caveats</title>
  <para>Here we use the post-build hook to upload to a binary cache.
  This is a simple and working example, but it is not suitable for all
  use cases.</para>

  <para>The post build hook program runs after each executed build,
  and blocks the build loop. The build loop exits if the hook program
  fails.</para>

  <para>Concretely, this implementation will make Nix slow or unusable
  when the internet is slow or unreliable.</para>

  <para>A more advanced implementation might pass the store paths to a
  user-supplied daemon or queue for processing the store paths outside
  of the build loop.</para>
</section>

<section>
  <title>Prerequisites</title>

  <para>
    This tutorial assumes you have configured an S3-compatible binary cache
    according to the instructions at
    <xref linkend="ssec-s3-substituter-authenticated-writes"/>, and
    that the <literal>root</literal> user's default AWS profile can
    upload to the bucket.
  </para>
</section>

<section>
  <title>Set up a Signing Key</title>
  <para>Use <command>nix-store --generate-binary-cache-key</command> to
  create our public and private signing keys. We will sign paths
  with the private key, and distribute the public key for verifying
  the authenticity of the paths.</para>

  <screen>
# nix-store --generate-binary-cache-key example-nix-cache-1 /etc/nix/key.private /etc/nix/key.public
# cat /etc/nix/key.public
example-nix-cache-1:1/cKDz3QCCOmwcztD2eV6Coggp6rqc9DGjWv7C0G+rM=
</screen>

<para>Then, add the public key and the cache URL to your
<filename>nix.conf</filename>'s <xref linkend="conf-trusted-public-keys"/>
and <xref linkend="conf-substituters"/> like:</para>

<programlisting>
substituters = https://cache.nixos.org/ s3://example-nix-cache
trusted-public-keys = cache.nixos.org-1:6NCHdD59X431o0gWypbMrAURkbJ16ZPMQFGspcDShjY= example-nix-cache-1:1/cKDz3QCCOmwcztD2eV6Coggp6rqc9DGjWv7C0G+rM=
</programlisting>

<para>We will restart the Nix daemon in a later step.</para>
</section>

<section>
  <title>Implementing the build hook</title>
  <para>Write the following script to
  <filename>/etc/nix/upload-to-cache.sh</filename>:
  </para>

  <programlisting>
#!/bin/sh

set -eu
set -f # disable globbing
export IFS=' '

echo "Signing paths" $OUT_PATHS
nix sign-paths --key-file /etc/nix/key.private $OUT_PATHS
echo "Uploading paths" $OUT_PATHS
exec nix copy --to 's3://example-nix-cache' $OUT_PATHS
</programlisting>

  <note>
    <title>Should <literal>$OUT_PATHS</literal> be quoted?</title>
    <para>
      The <literal>$OUT_PATHS</literal> variable is a space-separated
      list of Nix store paths. In this case, we expect and want the
      shell to perform word splitting to make each output path its
      own argument to <command>nix sign-paths</command>. Nix guarantees
      the paths will not contain any spaces, however a store path
      might contain glob characters. The <command>set -f</command>
      disables globbing in the shell.
    </para>
  </note>
  <para>
    Then make sure the hook program is executable by the <literal>root</literal> user:
    <screen>
# chmod +x /etc/nix/upload-to-cache.sh
</screen></para>
</section>

<section>
  <title>Updating Nix Configuration</title>

  <para>Edit <filename>/etc/nix/nix.conf</filename> to run our hook,
  by adding the following configuration snippet at the end:</para>

  <programlisting>
post-build-hook = /etc/nix/upload-to-cache.sh
</programlisting>

<para>Then, restart the <command>nix-daemon</command>.</para>
</section>

<section>
  <title>Testing</title>

  <para>Build any derivation, for example:</para>

  <screen>
$ nix-build -E '(import &lt;nixpkgs&gt; {}).writeText "example" (builtins.toString builtins.currentTime)'
this derivation will be built:
  /nix/store/s4pnfbkalzy5qz57qs6yybna8wylkig6-example.drv
building '/nix/store/s4pnfbkalzy5qz57qs6yybna8wylkig6-example.drv'...
running post-build-hook '/home/grahamc/projects/github.com/NixOS/nix/post-hook.sh'...
post-build-hook: Signing paths /nix/store/ibcyipq5gf91838ldx40mjsp0b8w9n18-example
post-build-hook: Uploading paths /nix/store/ibcyipq5gf91838ldx40mjsp0b8w9n18-example
/nix/store/ibcyipq5gf91838ldx40mjsp0b8w9n18-example
</screen>

  <para>Then delete the path from the store, and try substituting it from the binary cache:</para>
  <screen>
$ rm ./result
$ nix-store --delete /nix/store/ibcyipq5gf91838ldx40mjsp0b8w9n18-example
</screen>

<para>Now, copy the path back from the cache:</para>
<screen>
$ nix-store --realise /nix/store/ibcyipq5gf91838ldx40mjsp0b8w9n18-example
copying path '/nix/store/m8bmqwrch6l3h8s0k3d673xpmipcdpsa-example from 's3://example-nix-cache'...
warning: you did not specify '--add-root'; the result might be removed by the garbage collector
/nix/store/m8bmqwrch6l3h8s0k3d673xpmipcdpsa-example
</screen>
</section>
<section>
  <title>Conclusion</title>
  <para>
    We now have a Nix installation configured to automatically sign and
    upload every local build to a remote binary cache.
  </para>

  <para>
    Before deploying this to production, be sure to consider the
    implementation caveats in <xref linkend="chap-post-build-hook-caveats"/>.
  </para>
</section>
</chapter>

</part>
  <part xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="part-command-ref" xml:base="command-ref/command-ref.xml">

<title>Command Reference</title>

<partintro>
<para>This section lists commands and options that you can use when you
work with Nix.</para>
</partintro>

<chapter xmlns="http://docbook.org/ns/docbook" xml:id="sec-common-options">

<title>Common Options</title>


<para>Most Nix commands accept the following command-line options:</para>

<variablelist xml:id="opt-common">

<varlistentry><term><option>--help</option></term>

  <listitem><para>Prints out a summary of the command syntax and
  exits.</para></listitem>

</varlistentry>


<varlistentry><term><option>--version</option></term>

  <listitem><para>Prints out the Nix version number on standard output
  and exits.</para></listitem>
</varlistentry>


<varlistentry><term><option>--verbose</option> / <option>-v</option></term>

  <listitem>

  <para>Increases the level of verbosity of diagnostic messages
  printed on standard error.  For each Nix operation, the information
  printed on standard output is well-defined; any diagnostic
  information is printed on standard error, never on standard
  output.</para>

  <para>This option may be specified repeatedly.  Currently, the
  following verbosity levels exist:</para>

  <variablelist>

    <varlistentry><term>0</term>
    <listitem><para>&#x201C;Errors only&#x201D;: only print messages
    explaining why the Nix invocation failed.</para></listitem>
    </varlistentry>

    <varlistentry><term>1</term>
    <listitem><para>&#x201C;Informational&#x201D;: print
    <emphasis>useful</emphasis> messages about what Nix is doing.
    This is the default.</para></listitem>
    </varlistentry>

    <varlistentry><term>2</term>
    <listitem><para>&#x201C;Talkative&#x201D;: print more informational
    messages.</para></listitem>
    </varlistentry>

    <varlistentry><term>3</term>
    <listitem><para>&#x201C;Chatty&#x201D;: print even more
    informational messages.</para></listitem>
    </varlistentry>

    <varlistentry><term>4</term>
    <listitem><para>&#x201C;Debug&#x201D;: print debug
    information.</para></listitem>
    </varlistentry>

    <varlistentry><term>5</term>
    <listitem><para>&#x201C;Vomit&#x201D;: print vast amounts of debug
    information.</para></listitem>
    </varlistentry>

  </variablelist>

  </listitem>

</varlistentry>


<varlistentry><term><option>--quiet</option></term>

  <listitem>

  <para>Decreases the level of verbosity of diagnostic messages
  printed on standard error.  This is the inverse option to
  <option>-v</option> / <option>--verbose</option>.
  </para>

  <para>This option may be specified repeatedly.  See the previous
  verbosity levels list.</para>

  </listitem>

</varlistentry>


<varlistentry xml:id="opt-log-format"><term><option>--log-format</option> <replaceable>format</replaceable></term>

  <listitem>

  <para>This option can be used to change the output of the log format, with
  <replaceable>format</replaceable> being one of:</para>

  <variablelist>

    <varlistentry><term>raw</term>
    <listitem><para>This is the raw format, as outputted by nix-build.</para></listitem>
    </varlistentry>

    <varlistentry><term>internal-json</term>
    <listitem><para>Outputs the logs in a structured manner. NOTE: the json schema is not guarantees to be stable between releases.</para></listitem>
    </varlistentry>

    <varlistentry><term>bar</term>
    <listitem><para>Only display a progress bar during the builds.</para></listitem>
    </varlistentry>

    <varlistentry><term>bar-with-logs</term>
    <listitem><para>Display the raw logs, with the progress bar at the bottom.</para></listitem>
    </varlistentry>

  </variablelist>

  </listitem>

</varlistentry>

<varlistentry><term><option>--no-build-output</option> / <option>-Q</option></term>

  <listitem><para>By default, output written by builders to standard
  output and standard error is echoed to the Nix command's standard
  error.  This option suppresses this behaviour.  Note that the
  builder's standard output and error are always written to a log file
  in
  <filename><replaceable>prefix</replaceable>/nix/var/log/nix</filename>.</para></listitem>

</varlistentry>


<varlistentry xml:id="opt-max-jobs"><term><option>--max-jobs</option> / <option>-j</option>
<replaceable>number</replaceable></term>

  <listitem>

  <para>Sets the maximum number of build jobs that Nix will
  perform in parallel to the specified number.  Specify
  <literal>auto</literal> to use the number of CPUs in the system.
  The default is specified by the <link linkend="conf-max-jobs"><literal>max-jobs</literal></link>
  configuration setting, which itself defaults to
  <literal>1</literal>.  A higher value is useful on SMP systems or to
  exploit I/O latency.</para>

  <para> Setting it to <literal>0</literal> disallows building on the local
  machine, which is useful when you want builds to happen only on remote
  builders.</para>

  </listitem>

</varlistentry>


<varlistentry xml:id="opt-cores"><term><option>--cores</option></term>

  <listitem><para>Sets the value of the <envar>NIX_BUILD_CORES</envar>
  environment variable in the invocation of builders.  Builders can
  use this variable at their discretion to control the maximum amount
  of parallelism.  For instance, in Nixpkgs, if the derivation
  attribute <varname>enableParallelBuilding</varname> is set to
  <literal>true</literal>, the builder passes the
  <option>-j<replaceable>N</replaceable></option> flag to GNU Make.
  It defaults to the value of the <link linkend="conf-cores"><literal>cores</literal></link>
  configuration setting, if set, or <literal>1</literal> otherwise.
  The value <literal>0</literal> means that the builder should use all
  available CPU cores in the system.</para></listitem>

</varlistentry>


<varlistentry xml:id="opt-max-silent-time"><term><option>--max-silent-time</option></term>

  <listitem><para>Sets the maximum number of seconds that a builder
  can go without producing any data on standard output or standard
  error.  The default is specified by the <link linkend="conf-max-silent-time"><literal>max-silent-time</literal></link>
  configuration setting.  <literal>0</literal> means no
  time-out.</para></listitem>

</varlistentry>

<varlistentry xml:id="opt-timeout"><term><option>--timeout</option></term>

  <listitem><para>Sets the maximum number of seconds that a builder
  can run.  The default is specified by the <link linkend="conf-timeout"><literal>timeout</literal></link>
  configuration setting.  <literal>0</literal> means no
  timeout.</para></listitem>

</varlistentry>

<varlistentry><term><option>--keep-going</option> / <option>-k</option></term>

  <listitem><para>Keep going in case of failed builds, to the
  greatest extent possible.  That is, if building an input of some
  derivation fails, Nix will still build the other inputs, but not the
  derivation itself.  Without this option, Nix stops if any build
  fails (except for builds of substitutes), possibly killing builds in
  progress (in case of parallel or distributed builds).</para></listitem>

</varlistentry>


<varlistentry><term><option>--keep-failed</option> / <option>-K</option></term>

  <listitem><para>Specifies that in case of a build failure, the
  temporary directory (usually in <filename>/tmp</filename>) in which
  the build takes place should not be deleted.  The path of the build
  directory is printed as an informational message.
    </para>
  </listitem>
</varlistentry>


<varlistentry><term><option>--fallback</option></term>

  <listitem>

  <para>Whenever Nix attempts to build a derivation for which
  substitutes are known for each output path, but realising the output
  paths through the substitutes fails, fall back on building the
  derivation.</para>

  <para>The most common scenario in which this is useful is when we
  have registered substitutes in order to perform binary distribution
  from, say, a network repository.  If the repository is down, the
  realisation of the derivation will fail.  When this option is
  specified, Nix will build the derivation instead.  Thus,
  installation from binaries falls back on installation from source.
  This option is not the default since it is generally not desirable
  for a transient failure in obtaining the substitutes to lead to a
  full build from source (with the related consumption of
  resources).</para>

  </listitem>

</varlistentry>

<varlistentry><term><option>--no-build-hook</option></term>

  <listitem>

  <para>Disables the build hook mechanism.  This allows to ignore remote
  builders if they are setup on the machine.</para>

  <para>It's useful in cases where the bandwidth between the client and the
  remote builder is too low.  In that case it can take more time to upload the
  sources to the remote builder and fetch back the result than to do the
  computation locally.</para>

  </listitem>

</varlistentry>



<varlistentry><term><option>--readonly-mode</option></term>

  <listitem><para>When this option is used, no attempt is made to open
  the Nix database.  Most Nix operations do need database access, so
  those operations will fail.</para></listitem>

</varlistentry>


<varlistentry><term><option>--arg</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>This option is accepted by
  <command>nix-env</command>, <command>nix-instantiate</command>,
  <command>nix-shell</command> and <command>nix-build</command>.
  When evaluating Nix expressions, the expression evaluator will
  automatically try to call functions that
  it encounters.  It can automatically call functions for which every
  argument has a <link linkend="ss-functions">default value</link>
  (e.g., <literal>{ <replaceable>argName</replaceable> ?
  <replaceable>defaultValue</replaceable> }:
  <replaceable>...</replaceable></literal>).  With
  <option>--arg</option>, you can also call functions that have
  arguments without a default value (or override a default value).
  That is, if the evaluator encounters a function with an argument
  named <replaceable>name</replaceable>, it will call it with value
  <replaceable>value</replaceable>.</para>

  <para>For instance, the top-level <literal>default.nix</literal> in
  Nixpkgs is actually a function:

<programlisting>
{ # The system (e.g., `i686-linux') for which to build the packages.
  system ? builtins.currentSystem
  <replaceable>...</replaceable>
}: <replaceable>...</replaceable></programlisting>

  So if you call this Nix expression (e.g., when you do
  <literal>nix-env -i <replaceable>pkgname</replaceable></literal>),
  the function will be called automatically using the value <link linkend="builtin-currentSystem"><literal>builtins.currentSystem</literal></link>
  for the <literal>system</literal> argument.  You can override this
  using <option>--arg</option>, e.g., <literal>nix-env -i
  <replaceable>pkgname</replaceable> --arg system
  \"i686-freebsd\"</literal>.  (Note that since the argument is a Nix
  string literal, you have to escape the quotes.)</para></listitem>

</varlistentry>


<varlistentry><term><option>--argstr</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>This option is like <option>--arg</option>, only the
  value is not a Nix expression but a string.  So instead of
  <literal>--arg system \"i686-linux\"</literal> (the outer quotes are
  to keep the shell happy) you can say <literal>--argstr system
  i686-linux</literal>.</para></listitem>

</varlistentry>


<varlistentry xml:id="opt-attr"><term><option>--attr</option> / <option>-A</option>
<replaceable>attrPath</replaceable></term>

  <listitem><para>Select an attribute from the top-level Nix
  expression being evaluated.  (<command>nix-env</command>,
  <command>nix-instantiate</command>, <command>nix-build</command> and
  <command>nix-shell</command> only.)  The <emphasis>attribute
  path</emphasis> <replaceable>attrPath</replaceable> is a sequence of
  attribute names separated by dots.  For instance, given a top-level
  Nix expression <replaceable>e</replaceable>, the attribute path
  <literal>xorg.xorgserver</literal> would cause the expression
  <literal><replaceable>e</replaceable>.xorg.xorgserver</literal> to
  be used.  See <link linkend="refsec-nix-env-install-examples"><command>nix-env
  --install</command></link> for some concrete examples.</para>

  <para>In addition to attribute names, you can also specify array
  indices.  For instance, the attribute path
  <literal>foo.3.bar</literal> selects the <literal>bar</literal>
  attribute of the fourth element of the array in the
  <literal>foo</literal> attribute of the top-level
  expression.</para></listitem>

</varlistentry>


<varlistentry><term><option>--expr</option> / <option>-E</option></term>

  <listitem><para>Interpret the command line arguments as a list of
  Nix expressions to be parsed and evaluated, rather than as a list
  of file names of Nix expressions.
  (<command>nix-instantiate</command>, <command>nix-build</command>
  and <command>nix-shell</command> only.)</para>

  <para>For <command>nix-shell</command>, this option is commonly used
  to give you a shell in which you can build the packages returned
  by the expression. If you want to get a shell which contain the
  <emphasis>built</emphasis> packages ready for use, give your
  expression to the <command>nix-shell -p</command> convenience flag
  instead.</para></listitem>

</varlistentry>


<varlistentry xml:id="opt-I"><term><option>-I</option> <replaceable>path</replaceable></term>

  <listitem><para>Add a path to the Nix expression search path.  This
  option may be given multiple times.  See the <envar linkend="env-NIX_PATH">NIX_PATH</envar> environment variable for
  information on the semantics of the Nix search path.  Paths added
  through <option>-I</option> take precedence over
  <envar>NIX_PATH</envar>.</para></listitem>

</varlistentry>


<varlistentry><term><option>--option</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>Set the Nix configuration option
  <replaceable>name</replaceable> to <replaceable>value</replaceable>.
  This overrides settings in the Nix configuration file (see
  <citerefentry><refentrytitle>nix.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>).</para></listitem>

</varlistentry>


<varlistentry><term><option>--repair</option></term>

  <listitem><para>Fix corrupted or missing store paths by
  redownloading or rebuilding them.  Note that this is slow because it
  requires computing a cryptographic hash of the contents of every
  path in the closure of the build.  Also note the warning under
  <command>nix-store --repair-path</command>.</para></listitem>

</varlistentry>


</variablelist>


</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-common-env">

<title>Common Environment Variables</title>


<para>Most Nix commands interpret the following environment variables:</para>

<variablelist xml:id="env-common">

<varlistentry><term><envar>IN_NIX_SHELL</envar></term>

  <listitem><para>Indicator that tells if the current environment was set up by
  <command>nix-shell</command>.  Since Nix 2.0 the values are
  <literal>"pure"</literal> and <literal>"impure"</literal></para></listitem>

</varlistentry>

<varlistentry xml:id="env-NIX_PATH"><term><envar>NIX_PATH</envar></term>

  <listitem>

    <para>A colon-separated list of directories used to look up Nix
    expressions enclosed in angle brackets (i.e.,
    <literal>&lt;<replaceable>path</replaceable>&gt;</literal>).  For
    instance, the value

    <screen>
/home/eelco/Dev:/etc/nixos</screen>

    will cause Nix to look for paths relative to
    <filename>/home/eelco/Dev</filename> and
    <filename>/etc/nixos</filename>, in this order.  It is also
    possible to match paths against a prefix.  For example, the value

    <screen>
nixpkgs=/home/eelco/Dev/nixpkgs-branch:/etc/nixos</screen>

    will cause Nix to search for
    <literal>&lt;nixpkgs/<replaceable>path</replaceable>&gt;</literal> in
    <filename>/home/eelco/Dev/nixpkgs-branch/<replaceable>path</replaceable></filename>
    and
    <filename>/etc/nixos/nixpkgs/<replaceable>path</replaceable></filename>.</para>

    <para>If a path in the Nix search path starts with
    <literal>http://</literal> or <literal>https://</literal>, it is
    interpreted as the URL of a tarball that will be downloaded and
    unpacked to a temporary location. The tarball must consist of a
    single top-level directory. For example, setting
    <envar>NIX_PATH</envar> to

    <screen>
nixpkgs=https://github.com/NixOS/nixpkgs/archive/nixos-15.09.tar.gz</screen>

    tells Nix to download the latest revision in the Nixpkgs/NixOS
    15.09 channel.</para>

    <para>A following shorthand can be used to refer to the official channels:

    <screen>nixpkgs=channel:nixos-15.09</screen>
    </para>

    <para>The search path can be extended using the <option linkend="opt-I">-I</option> option, which takes precedence over
    <envar>NIX_PATH</envar>.</para></listitem>

</varlistentry>


<varlistentry><term><envar>NIX_IGNORE_SYMLINK_STORE</envar></term>

  <listitem>

  <para>Normally, the Nix store directory (typically
  <filename>/nix/store</filename>) is not allowed to contain any
  symlink components.  This is to prevent &#x201C;impure&#x201D; builds.  Builders
  sometimes &#x201C;canonicalise&#x201D; paths by resolving all symlink components.
  Thus, builds on different machines (with
  <filename>/nix/store</filename> resolving to different locations)
  could yield different results.  This is generally not a problem,
  except when builds are deployed to machines where
  <filename>/nix/store</filename> resolves differently.  If you are
  sure that you&#x2019;re not going to do that, you can set
  <envar>NIX_IGNORE_SYMLINK_STORE</envar> to <envar>1</envar>.</para>

  <para>Note that if you&#x2019;re symlinking the Nix store so that you can
  put it on another file system than the root file system, on Linux
  you&#x2019;re better off using <literal>bind</literal> mount points, e.g.,

  <screen>
$ mkdir /nix
$ mount -o bind /mnt/otherdisk/nix /nix</screen>

  Consult the <citerefentry><refentrytitle>mount</refentrytitle>
  <manvolnum>8</manvolnum></citerefentry> manual page for details.</para>

  </listitem>

</varlistentry>


<varlistentry><term><envar>NIX_STORE_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix store (default
  <filename><replaceable>prefix</replaceable>/store</filename>).</para></listitem>

</varlistentry>


<varlistentry><term><envar>NIX_DATA_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix static data
  directory (default
  <filename><replaceable>prefix</replaceable>/share</filename>).</para></listitem>

</varlistentry>


<varlistentry><term><envar>NIX_LOG_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix log directory
  (default <filename><replaceable>prefix</replaceable>/var/log/nix</filename>).</para></listitem>

</varlistentry>


<varlistentry><term><envar>NIX_STATE_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix state directory
  (default <filename><replaceable>prefix</replaceable>/var/nix</filename>).</para></listitem>

</varlistentry>


<varlistentry><term><envar>NIX_CONF_DIR</envar></term>

  <listitem><para>Overrides the location of the system Nix configuration
  directory (default
  <filename><replaceable>prefix</replaceable>/etc/nix</filename>).</para></listitem>

</varlistentry>

<varlistentry><term><envar>NIX_USER_CONF_FILES</envar></term>

  <listitem><para>Overrides the location of the user Nix configuration files
  to load from (defaults to the XDG spec locations). The variable is treated
  as a list separated by the <literal>:</literal> token.</para></listitem>

</varlistentry>

<varlistentry><term><envar>TMPDIR</envar></term>

  <listitem><para>Use the specified directory to store temporary
  files.  In particular, this includes temporary build directories;
  these can take up substantial amounts of disk space.  The default is
  <filename>/tmp</filename>.</para></listitem>

</varlistentry>


<varlistentry xml:id="envar-remote"><term><envar>NIX_REMOTE</envar></term>

  <listitem><para>This variable should be set to
  <literal>daemon</literal> if you want to use the Nix daemon to
  execute Nix operations. This is necessary in <link linkend="ssec-multi-user">multi-user Nix installations</link>.
  If the Nix daemon's Unix socket is at some non-standard path,
  this variable should be set to <literal>unix://path/to/socket</literal>.
  Otherwise, it should be left unset.</para></listitem>

</varlistentry>


<varlistentry><term><envar>NIX_SHOW_STATS</envar></term>

  <listitem><para>If set to <literal>1</literal>, Nix will print some
  evaluation statistics, such as the number of values
  allocated.</para></listitem>

</varlistentry>


<varlistentry><term><envar>NIX_COUNT_CALLS</envar></term>

  <listitem><para>If set to <literal>1</literal>, Nix will print how
  often functions were called during Nix expression evaluation.  This
  is useful for profiling your Nix expressions.</para></listitem>

</varlistentry>


<varlistentry><term><envar>GC_INITIAL_HEAP_SIZE</envar></term>

  <listitem><para>If Nix has been configured to use the Boehm garbage
  collector, this variable sets the initial size of the heap in bytes.
  It defaults to 384 MiB.  Setting it to a low value reduces memory
  consumption, but will increase runtime due to the overhead of
  garbage collection.</para></listitem>

</varlistentry>


</variablelist>


</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-main-commands">

<title>Main Commands</title>

<para>This section lists commands and options that you can use when you
work with Nix.</para>

<refentry xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-nix-env">

<refmeta>
  <refentrytitle>nix-env</refentrytitle>
  <manvolnum>1</manvolnum>
  <refmiscinfo class="source">Nix</refmiscinfo>
  <refmiscinfo class="version">3.0</refmiscinfo>
</refmeta>

<refnamediv>
  <refname>nix-env</refname>
  <refpurpose>manipulate or query Nix user environments</refpurpose>
</refnamediv>

<refsynopsisdiv>
  <cmdsynopsis>
    <command>nix-env</command>
    <arg xmlns="http://docbook.org/ns/docbook"><option>--help</option></arg><arg xmlns="http://docbook.org/ns/docbook"><option>--version</option></arg><arg xmlns="http://docbook.org/ns/docbook" rep="repeat">
  <group choice="req">
    <arg choice="plain"><option>--verbose</option></arg>
    <arg choice="plain"><option>-v</option></arg>
  </group>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <arg choice="plain"><option>--quiet</option></arg>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>--log-format</option>
  <replaceable>format</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <group choice="plain">
    <arg choice="plain"><option>--no-build-output</option></arg>
    <arg choice="plain"><option>-Q</option></arg>
  </group>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <group choice="req">
    <arg choice="plain"><option>--max-jobs</option></arg>
    <arg choice="plain"><option>-j</option></arg>
  </group>
  <replaceable>number</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>--cores</option>
  <replaceable>number</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>--max-silent-time</option>
  <replaceable>number</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>--timeout</option>
  <replaceable>number</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <group choice="plain">
    <arg choice="plain"><option>--keep-going</option></arg>
    <arg choice="plain"><option>-k</option></arg>
  </group>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <group choice="plain">
    <arg choice="plain"><option>--keep-failed</option></arg>
    <arg choice="plain"><option>-K</option></arg>
  </group>
</arg><arg xmlns="http://docbook.org/ns/docbook"><option>--fallback</option></arg><arg xmlns="http://docbook.org/ns/docbook"><option>--readonly-mode</option></arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>-I</option>
  <replaceable>path</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>--option</option>
  <replaceable>name</replaceable>
  <replaceable>value</replaceable>
</arg><sbr xmlns="http://docbook.org/ns/docbook"/>
    <arg><option>--arg</option> <replaceable>name</replaceable> <replaceable>value</replaceable></arg>
    <arg><option>--argstr</option> <replaceable>name</replaceable> <replaceable>value</replaceable></arg>
    <arg>
      <group choice="req">
        <arg choice="plain"><option>--file</option></arg>
        <arg choice="plain"><option>-f</option></arg>
      </group>
      <replaceable>path</replaceable>
    </arg>
    <arg>
      <group choice="req">
        <arg choice="plain"><option>--profile</option></arg>
        <arg choice="plain"><option>-p</option></arg>
      </group>
      <replaceable>path</replaceable>
    </arg>
    <arg>
      <arg choice="plain"><option>--system-filter</option></arg>
      <replaceable>system</replaceable>
    </arg>
    <arg><option>--dry-run</option></arg>
    <arg choice="plain"><replaceable>operation</replaceable></arg>
    <arg rep="repeat"><replaceable>options</replaceable></arg>
    <arg rep="repeat"><replaceable>arguments</replaceable></arg>
  </cmdsynopsis>
</refsynopsisdiv>


<refsection><title>Description</title>

<para>The command <command>nix-env</command> is used to manipulate Nix
user environments.  User environments are sets of software packages
available to a user at some point in time.  In other words, they are a
synthesised view of the programs available in the Nix store.  There
may be many user environments: different users can have different
environments, and individual users can switch between different
environments.</para>

<para><command>nix-env</command> takes exactly one
<emphasis>operation</emphasis> flag which indicates the subcommand to
be performed.  These are documented below.</para>

</refsection>



<!--######################################################################-->

<refsection><title>Selectors</title>

<para>Several commands, such as <command>nix-env -q</command> and
<command>nix-env -i</command>, take a list of arguments that specify
the packages on which to operate. These are extended regular
expressions that must match the entire name of the package. (For
details on regular expressions, see
<citerefentry><refentrytitle>regex</refentrytitle><manvolnum>7</manvolnum></citerefentry>.)
The match is case-sensitive. The regular expression can optionally be
followed by a dash and a version number; if omitted, any version of
the package will match.  Here are some examples:

<variablelist>

  <varlistentry>
    <term><literal>firefox</literal></term>
    <listitem><para>Matches the package name
    <literal>firefox</literal> and any version.</para></listitem>
  </varlistentry>

  <varlistentry>
    <term><literal>firefox-32.0</literal></term>
    <listitem><para>Matches the package name
    <literal>firefox</literal> and version
    <literal>32.0</literal>.</para></listitem>
  </varlistentry>

  <varlistentry>
    <term><literal>gtk\\+</literal></term>
    <listitem><para>Matches the package name
    <literal>gtk+</literal>. The <literal>+</literal> character must
    be escaped using a backslash to prevent it from being interpreted
    as a quantifier, and the backslash must be escaped in turn with
    another backslash to ensure that the shell passes it
    on.</para></listitem>
  </varlistentry>

  <varlistentry>
    <term><literal>.\*</literal></term>
    <listitem><para>Matches any package name. This is the default for
    most commands.</para></listitem>
  </varlistentry>

  <varlistentry>
    <term><literal>'.*zip.*'</literal></term>
    <listitem><para>Matches any package name containing the string
    <literal>zip</literal>. Note the dots: <literal>'*zip*'</literal>
    does not work, because in a regular expression, the character
    <literal>*</literal> is interpreted as a
    quantifier.</para></listitem>
  </varlistentry>

  <varlistentry>
    <term><literal>'.*(firefox|chromium).*'</literal></term>
    <listitem><para>Matches any package name containing the strings
    <literal>firefox</literal> or
    <literal>chromium</literal>.</para></listitem>
  </varlistentry>

</variablelist>

</para>

</refsection>



<!--######################################################################-->

<refsection><title>Common options</title>

<para>This section lists the options that are common to all
operations.  These options are allowed for every subcommand, though
they may not always have an effect.  <phrase condition="manual">See
also <xref linkend="sec-common-options"/>.</phrase></para>

<variablelist>

  <varlistentry><term><option>--file</option> / <option>-f</option> <replaceable>path</replaceable></term>

    <listitem><para>Specifies the Nix expression (designated below as
    the <emphasis>active Nix expression</emphasis>) used by the
    <option>--install</option>, <option>--upgrade</option>, and
    <option>--query --available</option> operations to obtain
    derivations.  The default is
    <filename>~/.nix-defexpr</filename>.</para>

    <para>If the argument starts with <literal>http://</literal> or
    <literal>https://</literal>, it is interpreted as the URL of a
    tarball that will be downloaded and unpacked to a temporary
    location. The tarball must include a single top-level directory
    containing at least a file named <filename>default.nix</filename>.</para>

    </listitem>

  </varlistentry>

  <varlistentry><term><option>--profile</option> / <option>-p</option> <replaceable>path</replaceable></term>

    <listitem><para>Specifies the profile to be used by those
    operations that operate on a profile (designated below as the
    <emphasis>active profile</emphasis>).  A profile is a sequence of
    user environments called <emphasis>generations</emphasis>, one of
    which is the <emphasis>current
    generation</emphasis>.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--dry-run</option></term>

    <listitem><para>For the <option>--install</option>,
    <option>--upgrade</option>, <option>--uninstall</option>,
    <option>--switch-generation</option>,
    <option>--delete-generations</option> and
    <option>--rollback</option> operations, this flag will cause
    <command>nix-env</command> to print what
    <emphasis>would</emphasis> be done if this flag had not been
    specified, without actually doing it.</para>

    <para><option>--dry-run</option> also prints out which paths will
    be <link linkend="gloss-substitute">substituted</link> (i.e.,
    downloaded) and which paths will be built from source (because no
    substitute is available).</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--system-filter</option> <replaceable>system</replaceable></term>

    <listitem><para>By default, operations such as <option>--query
    --available</option> show derivations matching any platform.  This
    option allows you to use derivations for the specified platform
    <replaceable>system</replaceable>.</para></listitem>

  </varlistentry>

</variablelist>

<variablelist condition="manpage">
  <varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--help</option></term>

  <listitem><para>Prints out a summary of the command syntax and
  exits.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--version</option></term>

  <listitem><para>Prints out the Nix version number on standard output
  and exits.</para></listitem>
</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--verbose</option> / <option>-v</option></term>

  <listitem>

  <para>Increases the level of verbosity of diagnostic messages
  printed on standard error.  For each Nix operation, the information
  printed on standard output is well-defined; any diagnostic
  information is printed on standard error, never on standard
  output.</para>

  <para>This option may be specified repeatedly.  Currently, the
  following verbosity levels exist:</para>

  <variablelist>

    <varlistentry><term>0</term>
    <listitem><para>&#x201C;Errors only&#x201D;: only print messages
    explaining why the Nix invocation failed.</para></listitem>
    </varlistentry>

    <varlistentry><term>1</term>
    <listitem><para>&#x201C;Informational&#x201D;: print
    <emphasis>useful</emphasis> messages about what Nix is doing.
    This is the default.</para></listitem>
    </varlistentry>

    <varlistentry><term>2</term>
    <listitem><para>&#x201C;Talkative&#x201D;: print more informational
    messages.</para></listitem>
    </varlistentry>

    <varlistentry><term>3</term>
    <listitem><para>&#x201C;Chatty&#x201D;: print even more
    informational messages.</para></listitem>
    </varlistentry>

    <varlistentry><term>4</term>
    <listitem><para>&#x201C;Debug&#x201D;: print debug
    information.</para></listitem>
    </varlistentry>

    <varlistentry><term>5</term>
    <listitem><para>&#x201C;Vomit&#x201D;: print vast amounts of debug
    information.</para></listitem>
    </varlistentry>

  </variablelist>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--quiet</option></term>

  <listitem>

  <para>Decreases the level of verbosity of diagnostic messages
  printed on standard error.  This is the inverse option to
  <option>-v</option> / <option>--verbose</option>.
  </para>

  <para>This option may be specified repeatedly.  See the previous
  verbosity levels list.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-log-format"><term><option>--log-format</option> <replaceable>format</replaceable></term>

  <listitem>

  <para>This option can be used to change the output of the log format, with
  <replaceable>format</replaceable> being one of:</para>

  <variablelist>

    <varlistentry><term>raw</term>
    <listitem><para>This is the raw format, as outputted by nix-build.</para></listitem>
    </varlistentry>

    <varlistentry><term>internal-json</term>
    <listitem><para>Outputs the logs in a structured manner. NOTE: the json schema is not guarantees to be stable between releases.</para></listitem>
    </varlistentry>

    <varlistentry><term>bar</term>
    <listitem><para>Only display a progress bar during the builds.</para></listitem>
    </varlistentry>

    <varlistentry><term>bar-with-logs</term>
    <listitem><para>Display the raw logs, with the progress bar at the bottom.</para></listitem>
    </varlistentry>

  </variablelist>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--no-build-output</option> / <option>-Q</option></term>

  <listitem><para>By default, output written by builders to standard
  output and standard error is echoed to the Nix command's standard
  error.  This option suppresses this behaviour.  Note that the
  builder's standard output and error are always written to a log file
  in
  <filename><replaceable>prefix</replaceable>/nix/var/log/nix</filename>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-max-jobs"><term><option>--max-jobs</option> / <option>-j</option>
<replaceable>number</replaceable></term>

  <listitem>

  <para>Sets the maximum number of build jobs that Nix will
  perform in parallel to the specified number.  Specify
  <literal>auto</literal> to use the number of CPUs in the system.
  The default is specified by the <link linkend="conf-max-jobs"><literal>max-jobs</literal></link>
  configuration setting, which itself defaults to
  <literal>1</literal>.  A higher value is useful on SMP systems or to
  exploit I/O latency.</para>

  <para> Setting it to <literal>0</literal> disallows building on the local
  machine, which is useful when you want builds to happen only on remote
  builders.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-cores"><term><option>--cores</option></term>

  <listitem><para>Sets the value of the <envar>NIX_BUILD_CORES</envar>
  environment variable in the invocation of builders.  Builders can
  use this variable at their discretion to control the maximum amount
  of parallelism.  For instance, in Nixpkgs, if the derivation
  attribute <varname>enableParallelBuilding</varname> is set to
  <literal>true</literal>, the builder passes the
  <option>-j<replaceable>N</replaceable></option> flag to GNU Make.
  It defaults to the value of the <link linkend="conf-cores"><literal>cores</literal></link>
  configuration setting, if set, or <literal>1</literal> otherwise.
  The value <literal>0</literal> means that the builder should use all
  available CPU cores in the system.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-max-silent-time"><term><option>--max-silent-time</option></term>

  <listitem><para>Sets the maximum number of seconds that a builder
  can go without producing any data on standard output or standard
  error.  The default is specified by the <link linkend="conf-max-silent-time"><literal>max-silent-time</literal></link>
  configuration setting.  <literal>0</literal> means no
  time-out.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-timeout"><term><option>--timeout</option></term>

  <listitem><para>Sets the maximum number of seconds that a builder
  can run.  The default is specified by the <link linkend="conf-timeout"><literal>timeout</literal></link>
  configuration setting.  <literal>0</literal> means no
  timeout.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--keep-going</option> / <option>-k</option></term>

  <listitem><para>Keep going in case of failed builds, to the
  greatest extent possible.  That is, if building an input of some
  derivation fails, Nix will still build the other inputs, but not the
  derivation itself.  Without this option, Nix stops if any build
  fails (except for builds of substitutes), possibly killing builds in
  progress (in case of parallel or distributed builds).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--keep-failed</option> / <option>-K</option></term>

  <listitem><para>Specifies that in case of a build failure, the
  temporary directory (usually in <filename>/tmp</filename>) in which
  the build takes place should not be deleted.  The path of the build
  directory is printed as an informational message.
    </para>
  </listitem>
</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--fallback</option></term>

  <listitem>

  <para>Whenever Nix attempts to build a derivation for which
  substitutes are known for each output path, but realising the output
  paths through the substitutes fails, fall back on building the
  derivation.</para>

  <para>The most common scenario in which this is useful is when we
  have registered substitutes in order to perform binary distribution
  from, say, a network repository.  If the repository is down, the
  realisation of the derivation will fail.  When this option is
  specified, Nix will build the derivation instead.  Thus,
  installation from binaries falls back on installation from source.
  This option is not the default since it is generally not desirable
  for a transient failure in obtaining the substitutes to lead to a
  full build from source (with the related consumption of
  resources).</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--no-build-hook</option></term>

  <listitem>

  <para>Disables the build hook mechanism.  This allows to ignore remote
  builders if they are setup on the machine.</para>

  <para>It's useful in cases where the bandwidth between the client and the
  remote builder is too low.  In that case it can take more time to upload the
  sources to the remote builder and fetch back the result than to do the
  computation locally.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--readonly-mode</option></term>

  <listitem><para>When this option is used, no attempt is made to open
  the Nix database.  Most Nix operations do need database access, so
  those operations will fail.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--arg</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>This option is accepted by
  <command>nix-env</command>, <command>nix-instantiate</command>,
  <command>nix-shell</command> and <command>nix-build</command>.
  When evaluating Nix expressions, the expression evaluator will
  automatically try to call functions that
  it encounters.  It can automatically call functions for which every
  argument has a <link linkend="ss-functions">default value</link>
  (e.g., <literal>{ <replaceable>argName</replaceable> ?
  <replaceable>defaultValue</replaceable> }:
  <replaceable>...</replaceable></literal>).  With
  <option>--arg</option>, you can also call functions that have
  arguments without a default value (or override a default value).
  That is, if the evaluator encounters a function with an argument
  named <replaceable>name</replaceable>, it will call it with value
  <replaceable>value</replaceable>.</para>

  <para>For instance, the top-level <literal>default.nix</literal> in
  Nixpkgs is actually a function:

<programlisting>
{ # The system (e.g., `i686-linux') for which to build the packages.
  system ? builtins.currentSystem
  <replaceable>...</replaceable>
}: <replaceable>...</replaceable></programlisting>

  So if you call this Nix expression (e.g., when you do
  <literal>nix-env -i <replaceable>pkgname</replaceable></literal>),
  the function will be called automatically using the value <link linkend="builtin-currentSystem"><literal>builtins.currentSystem</literal></link>
  for the <literal>system</literal> argument.  You can override this
  using <option>--arg</option>, e.g., <literal>nix-env -i
  <replaceable>pkgname</replaceable> --arg system
  \"i686-freebsd\"</literal>.  (Note that since the argument is a Nix
  string literal, you have to escape the quotes.)</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--argstr</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>This option is like <option>--arg</option>, only the
  value is not a Nix expression but a string.  So instead of
  <literal>--arg system \"i686-linux\"</literal> (the outer quotes are
  to keep the shell happy) you can say <literal>--argstr system
  i686-linux</literal>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-attr"><term><option>--attr</option> / <option>-A</option>
<replaceable>attrPath</replaceable></term>

  <listitem><para>Select an attribute from the top-level Nix
  expression being evaluated.  (<command>nix-env</command>,
  <command>nix-instantiate</command>, <command>nix-build</command> and
  <command>nix-shell</command> only.)  The <emphasis>attribute
  path</emphasis> <replaceable>attrPath</replaceable> is a sequence of
  attribute names separated by dots.  For instance, given a top-level
  Nix expression <replaceable>e</replaceable>, the attribute path
  <literal>xorg.xorgserver</literal> would cause the expression
  <literal><replaceable>e</replaceable>.xorg.xorgserver</literal> to
  be used.  See <link linkend="refsec-nix-env-install-examples"><command>nix-env
  --install</command></link> for some concrete examples.</para>

  <para>In addition to attribute names, you can also specify array
  indices.  For instance, the attribute path
  <literal>foo.3.bar</literal> selects the <literal>bar</literal>
  attribute of the fourth element of the array in the
  <literal>foo</literal> attribute of the top-level
  expression.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--expr</option> / <option>-E</option></term>

  <listitem><para>Interpret the command line arguments as a list of
  Nix expressions to be parsed and evaluated, rather than as a list
  of file names of Nix expressions.
  (<command>nix-instantiate</command>, <command>nix-build</command>
  and <command>nix-shell</command> only.)</para>

  <para>For <command>nix-shell</command>, this option is commonly used
  to give you a shell in which you can build the packages returned
  by the expression. If you want to get a shell which contain the
  <emphasis>built</emphasis> packages ready for use, give your
  expression to the <command>nix-shell -p</command> convenience flag
  instead.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-I"><term><option>-I</option> <replaceable>path</replaceable></term>

  <listitem><para>Add a path to the Nix expression search path.  This
  option may be given multiple times.  See the <envar linkend="env-NIX_PATH">NIX_PATH</envar> environment variable for
  information on the semantics of the Nix search path.  Paths added
  through <option>-I</option> take precedence over
  <envar>NIX_PATH</envar>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--option</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>Set the Nix configuration option
  <replaceable>name</replaceable> to <replaceable>value</replaceable>.
  This overrides settings in the Nix configuration file (see
  <citerefentry><refentrytitle>nix.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--repair</option></term>

  <listitem><para>Fix corrupted or missing store paths by
  redownloading or rebuilding them.  Note that this is slow because it
  requires computing a cryptographic hash of the contents of every
  path in the closure of the build.  Also note the warning under
  <command>nix-store --repair-path</command>.</para></listitem>

</varlistentry>
</variablelist>

</refsection>



<!--######################################################################-->

<refsection><title>Files</title>

<variablelist>

  <varlistentry><term><filename>~/.nix-defexpr</filename></term>

    <listitem><para>The source for the default Nix
    expressions used by the <option>--install</option>,
    <option>--upgrade</option>, and <option>--query
    --available</option> operations to obtain derivations. The
    <option>--file</option> option may be used to override this
    default.</para>

    <para>If <filename>~/.nix-defexpr</filename> is a file,
    it is loaded as a Nix expression. If the expression
    is a set, it is used as the default Nix expression.
    If the expression is a function, an empty set is passed
    as argument and the return value is used as
    the default Nix expression.</para>

    <para>If <filename>~/.nix-defexpr</filename> is a directory
    containing a <filename>default.nix</filename> file, that file
    is loaded as in the above paragraph.</para>

    <para>If <filename>~/.nix-defexpr</filename> is a directory without
    a <filename>default.nix</filename> file, then its contents
    (both files and subdirectories) are loaded as Nix expressions.
    The expressions are combined into a single set, each expression
    under an attribute with the same name as the original file
    or subdirectory.
    </para>

    <para>For example, if <filename>~/.nix-defexpr</filename> contains
    two files, <filename>foo.nix</filename> and <filename>bar.nix</filename>,
    then the default Nix expression will essentially be

<programlisting>
{
  foo = import ~/.nix-defexpr/foo.nix;
  bar = import ~/.nix-defexpr/bar.nix;
}</programlisting>

    </para>

    <para>The file <filename>manifest.nix</filename> is always ignored.
    Subdirectories without a <filename>default.nix</filename> file
    are traversed recursively in search of more Nix expressions,
    but the names of these intermediate directories are not
    added to the attribute paths of the default Nix expression.</para>

    <para>The command <command>nix-channel</command> places symlinks
    to the downloaded Nix expressions from each subscribed channel in
    this directory.</para>
    </listitem>

  </varlistentry>

  <varlistentry><term><filename>~/.nix-profile</filename></term>

    <listitem><para>A symbolic link to the user's current profile.  By
    default, this symlink points to
    <filename><replaceable>prefix</replaceable>/var/nix/profiles/default</filename>.
    The <envar>PATH</envar> environment variable should include
    <filename>~/.nix-profile/bin</filename> for the user environment
    to be visible to the user.</para></listitem>

  </varlistentry>

</variablelist>

</refsection>



<!--######################################################################-->

<refsection xml:id="rsec-nix-env-install"><title>Operation <option>--install</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-env</command>
  <group choice="req">
    <arg choice="plain"><option>--install</option></arg>
    <arg choice="plain"><option>-i</option></arg>
  </group>
  <arg xmlns="http://docbook.org/ns/docbook">
    <group choice="req">
      <arg choice="plain"><option>--prebuilt-only</option></arg>
      <arg choice="plain"><option>-b</option></arg>
    </group>
  </arg><arg xmlns="http://docbook.org/ns/docbook">
    <group choice="req">
      <arg choice="plain"><option>--attr</option></arg>
      <arg choice="plain"><option>-A</option></arg>
    </group>
  </arg><arg xmlns="http://docbook.org/ns/docbook"><option>--from-expression</option></arg><arg xmlns="http://docbook.org/ns/docbook"><option>-E</option></arg><arg xmlns="http://docbook.org/ns/docbook"><option>--from-profile</option> <replaceable>path</replaceable></arg>
  <group choice="opt">
    <arg choice="plain"><option>--preserve-installed</option></arg>
    <arg choice="plain"><option>-P</option></arg>
  </group>
  <group choice="opt">
    <arg choice="plain"><option>--remove-all</option></arg>
    <arg choice="plain"><option>-r</option></arg>
  </group>
  <arg choice="plain" rep="repeat"><replaceable>args</replaceable></arg>
</cmdsynopsis>

</refsection>


<refsection><title>Description</title>

<para>The install operation creates a new user environment, based on
the current generation of the active profile, to which a set of store
paths described by <replaceable>args</replaceable> is added.  The
arguments <replaceable>args</replaceable> map to store paths in a
number of possible ways:

<itemizedlist>

  <listitem><para>By default, <replaceable>args</replaceable> is a set
  of derivation names denoting derivations in the active Nix
  expression.  These are realised, and the resulting output paths are
  installed.  Currently installed derivations with a name equal to the
  name of a derivation being added are removed unless the option
  <option>--preserve-installed</option> is
  specified.</para>

  <para>If there are multiple derivations matching a name in
  <replaceable>args</replaceable> that have the same name (e.g.,
  <literal>gcc-3.3.6</literal> and <literal>gcc-4.1.1</literal>), then
  the derivation with the highest <emphasis>priority</emphasis> is
  used.  A derivation can define a priority by declaring the
  <varname>meta.priority</varname> attribute.  This attribute should
  be a number, with a higher value denoting a lower priority.  The
  default priority is <literal>0</literal>.</para>

  <para>If there are multiple matching derivations with the same
  priority, then the derivation with the highest version will be
  installed.</para>

  <para>You can force the installation of multiple derivations with
  the same name by being specific about the versions.  For instance,
  <literal>nix-env -i gcc-3.3.6 gcc-4.1.1</literal> will install both
  version of GCC (and will probably cause a user environment
  conflict!).</para></listitem>

  <listitem><para>If <link linkend="opt-attr"><option>--attr</option></link>
  (<option>-A</option>) is specified, the arguments are
  <emphasis>attribute paths</emphasis> that select attributes from the
  top-level Nix expression.  This is faster than using derivation
  names and unambiguous.  To find out the attribute paths of available
  packages, use <literal>nix-env -qaP</literal>.</para></listitem>

  <listitem><para>If <option>--from-profile</option>
  <replaceable>path</replaceable> is given,
  <replaceable>args</replaceable> is a set of names denoting installed
  store paths in the profile <replaceable>path</replaceable>.  This is
  an easy way to copy user environment elements from one profile to
  another.</para></listitem>

  <listitem><para>If <option>--from-expression</option> is given,
  <replaceable>args</replaceable> are Nix <link linkend="ss-functions">functions</link> that are called with the
  active Nix expression as their single argument.  The derivations
  returned by those function calls are installed.  This allows
  derivations to be specified in an unambiguous way, which is necessary
  if there are multiple derivations with the same
  name.</para></listitem>

  <listitem><para>If <replaceable>args</replaceable> are store
  derivations, then these are <link linkend="rsec-nix-store-realise">realised</link>, and the resulting
  output paths are installed.</para></listitem>

  <listitem><para>If <replaceable>args</replaceable> are store paths
  that are not store derivations, then these are <link linkend="rsec-nix-store-realise">realised</link> and
  installed.</para></listitem>

  <listitem><para>By default all outputs are installed for each derivation.
  That can be reduced by setting <literal>meta.outputsToInstall</literal>.
  </para></listitem> <!-- TODO: link nixpkgs docs on the ability to override those. -->

</itemizedlist>

</para>

</refsection>


<refsection><title>Flags</title>

<variablelist>

  <varlistentry><term><option>--prebuilt-only</option> / <option>-b</option></term>

    <listitem><para>Use only derivations for which a substitute is
    registered, i.e., there is a pre-built binary available that can
    be downloaded in lieu of building the derivation.  Thus, no
    packages will be built from source.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--preserve-installed</option></term>
    <term><option>-P</option></term>

    <listitem><para>Do not remove derivations with a name matching one
    of the derivations being installed.  Usually, trying to have two
    versions of the same package installed in the same generation of a
    profile will lead to an error in building the generation, due to
    file name clashes between the two versions.  However, this is not
    the case for all packages.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--remove-all</option></term>
    <term><option>-r</option></term>

    <listitem><para>Remove all previously installed packages first.
    This is equivalent to running <literal>nix-env -e '.*'</literal>
    first, except that everything happens in a single
    transaction.</para></listitem>

  </varlistentry>

</variablelist>

</refsection>


<refsection xml:id="refsec-nix-env-install-examples"><title>Examples</title>

<para>To install a specific version of <command>gcc</command> from the
active Nix expression:

<screen>
$ nix-env --install gcc-3.3.2
installing `gcc-3.3.2'
uninstalling `gcc-3.1'</screen>

Note the previously installed version is removed, since
<option>--preserve-installed</option> was not specified.</para>

<para>To install an arbitrary version:

<screen>
$ nix-env --install gcc
installing `gcc-3.3.2'</screen>

</para>

<para>To install using a specific attribute:

<screen>
$ nix-env -i -A gcc40mips
$ nix-env -i -A xorg.xorgserver</screen>

</para>

<para>To install all derivations in the Nix expression <filename>foo.nix</filename>:

<screen>
$ nix-env -f ~/foo.nix -i '.*'</screen>

</para>

<para>To copy the store path with symbolic name <literal>gcc</literal>
from another profile:

<screen>
$ nix-env -i --from-profile /nix/var/nix/profiles/foo gcc</screen>

</para>

<para>To install a specific store derivation (typically created by
<command>nix-instantiate</command>):

<screen>
$ nix-env -i /nix/store/fibjb1bfbpm5mrsxc4mh2d8n37sxh91i-gcc-3.4.3.drv</screen>

</para>

<para>To install a specific output path:

<screen>
$ nix-env -i /nix/store/y3cgx0xj1p4iv9x0pnnmdhr8iyg741vk-gcc-3.4.3</screen>

</para>

<para>To install from a Nix expression specified on the command-line:

<screen>
$ nix-env -f ./foo.nix -i -E \
    'f: (f {system = "i686-linux";}).subversionWithJava'</screen>

I.e., this evaluates to <literal>(f: (f {system =
"i686-linux";}).subversionWithJava) (import ./foo.nix)</literal>, thus
selecting the <literal>subversionWithJava</literal> attribute from the
set returned by calling the function defined in
<filename>./foo.nix</filename>.</para>

<para>A dry-run tells you which paths will be downloaded or built from
source:

<screen>
$ nix-env -f '&lt;nixpkgs&gt;' -iA hello --dry-run
(dry run; not doing anything)
installing &#x2018;hello-2.10&#x2019;
this path will be fetched (0.04 MiB download, 0.19 MiB unpacked):
  /nix/store/wkhdf9jinag5750mqlax6z2zbwhqb76n-hello-2.10
  <replaceable>...</replaceable></screen>

</para>

<para>To install Firefox from the latest revision in the Nixpkgs/NixOS
14.12 channel:

<screen>
$ nix-env -f https://github.com/NixOS/nixpkgs/archive/nixos-14.12.tar.gz -iA firefox
</screen>

</para>

</refsection>

</refsection>



<!--######################################################################-->

<refsection xml:id="rsec-nix-env-upgrade"><title>Operation <option>--upgrade</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-env</command>
  <group choice="req">
    <arg choice="plain"><option>--upgrade</option></arg>
    <arg choice="plain"><option>-u</option></arg>
  </group>
  <arg xmlns="http://docbook.org/ns/docbook">
    <group choice="req">
      <arg choice="plain"><option>--prebuilt-only</option></arg>
      <arg choice="plain"><option>-b</option></arg>
    </group>
  </arg><arg xmlns="http://docbook.org/ns/docbook">
    <group choice="req">
      <arg choice="plain"><option>--attr</option></arg>
      <arg choice="plain"><option>-A</option></arg>
    </group>
  </arg><arg xmlns="http://docbook.org/ns/docbook"><option>--from-expression</option></arg><arg xmlns="http://docbook.org/ns/docbook"><option>-E</option></arg><arg xmlns="http://docbook.org/ns/docbook"><option>--from-profile</option> <replaceable>path</replaceable></arg>
  <group choice="opt">
    <arg choice="plain"><option>--lt</option></arg>
    <arg choice="plain"><option>--leq</option></arg>
    <arg choice="plain"><option>--eq</option></arg>
    <arg choice="plain"><option>--always</option></arg>
  </group>
  <arg choice="plain" rep="repeat"><replaceable>args</replaceable></arg>
</cmdsynopsis>

</refsection>

<refsection><title>Description</title>

<para>The upgrade operation creates a new user environment, based on
the current generation of the active profile, in which all store paths
are replaced for which there are newer versions in the set of paths
described by <replaceable>args</replaceable>.  Paths for which there
are no newer versions are left untouched; this is not an error.  It is
also not an error if an element of <replaceable>args</replaceable>
matches no installed derivations.</para>

<para>For a description of how <replaceable>args</replaceable> is
mapped to a set of store paths, see <link linkend="rsec-nix-env-install"><option>--install</option></link>.  If
<replaceable>args</replaceable> describes multiple store paths with
the same symbolic name, only the one with the highest version is
installed.</para>

</refsection>

<refsection><title>Flags</title>

<variablelist>

  <varlistentry><term><option>--lt</option></term>

    <listitem><para>Only upgrade a derivation to newer versions.  This
    is the default.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--leq</option></term>

    <listitem><para>In addition to upgrading to newer versions, also
    &#x201C;upgrade&#x201D; to derivations that have the same version.  Version are
    not a unique identification of a derivation, so there may be many
    derivations that have the same version.  This flag may be useful
    to force &#x201C;synchronisation&#x201D; between the installed and available
    derivations.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--eq</option></term>

    <listitem><para><emphasis>Only</emphasis> &#x201C;upgrade&#x201D; to derivations
    that have the same version.  This may not seem very useful, but it
    actually is, e.g., when there is a new release of Nixpkgs and you
    want to replace installed applications with the same versions
    built against newer dependencies (to reduce the number of
    dependencies floating around on your system).</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--always</option></term>

    <listitem><para>In addition to upgrading to newer versions, also
    &#x201C;upgrade&#x201D; to derivations that have the same or a lower version.
    I.e., derivations may actually be downgraded depending on what is
    available in the active Nix expression.</para></listitem>

  </varlistentry>

</variablelist>

<para>For the other flags, see <option linkend="rsec-nix-env-install">--install</option>.</para>

</refsection>

<refsection><title>Examples</title>

<screen>
$ nix-env --upgrade gcc
upgrading `gcc-3.3.1' to `gcc-3.4'

$ nix-env -u gcc-3.3.2 --always <lineannotation>(switch to a specific version)</lineannotation>
upgrading `gcc-3.4' to `gcc-3.3.2'

$ nix-env --upgrade pan
<lineannotation>(no upgrades available, so nothing happens)</lineannotation>

$ nix-env -u <lineannotation>(try to upgrade everything)</lineannotation>
upgrading `hello-2.1.2' to `hello-2.1.3'
upgrading `mozilla-1.2' to `mozilla-1.4'</screen>

</refsection>

<refsection xml:id="ssec-version-comparisons"><title>Versions</title>

<para>The upgrade operation determines whether a derivation
<varname>y</varname> is an upgrade of a derivation
<varname>x</varname> by looking at their respective
<literal>name</literal> attributes.  The names (e.g.,
<literal>gcc-3.3.1</literal> are split into two parts: the package
name (<literal>gcc</literal>), and the version
(<literal>3.3.1</literal>).  The version part starts after the first
dash not followed by a letter.  <varname>x</varname> is considered an
upgrade of <varname>y</varname> if their package names match, and the
version of <varname>y</varname> is higher that that of
<varname>x</varname>.</para>

<para>The versions are compared by splitting them into contiguous
components of numbers and letters.  E.g., <literal>3.3.1pre5</literal>
is split into <literal>[3, 3, 1, "pre", 5]</literal>.  These lists are
then compared lexicographically (from left to right).  Corresponding
components <varname>a</varname> and <varname>b</varname> are compared
as follows.  If they are both numbers, integer comparison is used.  If
<varname>a</varname> is an empty string and <varname>b</varname> is a
number, <varname>a</varname> is considered less than
<varname>b</varname>.  The special string component
<literal>pre</literal> (for <emphasis>pre-release</emphasis>) is
considered to be less than other components.  String components are
considered less than number components.  Otherwise, they are compared
lexicographically (i.e., using case-sensitive string comparison).</para>

<para>This is illustrated by the following examples:

<screen>
1.0 &lt; 2.3
2.1 &lt; 2.3
2.3 = 2.3
2.5 &gt; 2.3
3.1 &gt; 2.3
2.3.1 &gt; 2.3
2.3.1 &gt; 2.3a
2.3pre1 &lt; 2.3
2.3pre3 &lt; 2.3pre12
2.3a &lt; 2.3c
2.3pre1 &lt; 2.3c
2.3pre1 &lt; 2.3q</screen>

</para>

</refsection>

</refsection>



<!--######################################################################-->

<refsection><title>Operation <option>--uninstall</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-env</command>
  <group choice="req">
    <arg choice="plain"><option>--uninstall</option></arg>
    <arg choice="plain"><option>-e</option></arg>
  </group>
  <arg choice="plain" rep="repeat"><replaceable>drvnames</replaceable></arg>
</cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>The uninstall operation creates a new user environment, based on
the current generation of the active profile, from which the store
paths designated by the symbolic names
<replaceable>names</replaceable> are removed.</para>

</refsection>

<refsection><title>Examples</title>

<screen>
$ nix-env --uninstall gcc
$ nix-env -e '.*' <lineannotation>(remove everything)</lineannotation></screen>

</refsection>

</refsection>



<!--######################################################################-->

<refsection xml:id="rsec-nix-env-set"><title>Operation <option>--set</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-env</command>
  <arg choice="plain"><option>--set</option></arg>
  <arg choice="plain"><replaceable>drvname</replaceable></arg>
</cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>The <option>--set</option> operation modifies the current generation of a
profile so that it contains exactly the specified derivation, and nothing else.
</para>

</refsection>

<refsection><title>Examples</title>

<para>
The following updates a profile such that its current generation will contain
just Firefox:

<screen>
$ nix-env -p /nix/var/nix/profiles/browser --set firefox</screen>

</para>

</refsection>

</refsection>



<!--######################################################################-->

<refsection xml:id="rsec-nix-env-set-flag"><title>Operation <option>--set-flag</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-env</command>
  <arg choice="plain"><option>--set-flag</option></arg>
  <arg choice="plain"><replaceable>name</replaceable></arg>
  <arg choice="plain"><replaceable>value</replaceable></arg>
  <arg choice="plain" rep="repeat"><replaceable>drvnames</replaceable></arg>
</cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>The <option>--set-flag</option> operation allows meta attributes
of installed packages to be modified.  There are several attributes
that can be usefully modified, because they affect the behaviour of
<command>nix-env</command> or the user environment build
script:

<itemizedlist>

  <listitem><para><varname>priority</varname> can be changed to
  resolve filename clashes.  The user environment build script uses
  the <varname>meta.priority</varname> attribute of derivations to
  resolve filename collisions between packages.  Lower priority values
  denote a higher priority.  For instance, the GCC wrapper package and
  the Binutils package in Nixpkgs both have a file
  <filename>bin/ld</filename>, so previously if you tried to install
  both you would get a collision.  Now, on the other hand, the GCC
  wrapper declares a higher priority than Binutils, so the former&#x2019;s
  <filename>bin/ld</filename> is symlinked in the user
  environment.</para></listitem>

  <listitem><para><varname>keep</varname> can be set to
  <literal>true</literal> to prevent the package from being upgraded
  or replaced.  This is useful if you want to hang on to an older
  version of a package.</para></listitem>

  <listitem><para><varname>active</varname> can be set to
  <literal>false</literal> to &#x201C;disable&#x201D; the package.  That is, no
  symlinks will be generated to the files of the package, but it
  remains part of the profile (so it won&#x2019;t be garbage-collected).  It
  can be set back to <literal>true</literal> to re-enable the
  package.</para></listitem>

</itemizedlist>

</para>

</refsection>

<refsection><title>Examples</title>

<para>To prevent the currently installed Firefox from being upgraded:

<screen>
$ nix-env --set-flag keep true firefox</screen>

After this, <command>nix-env -u</command> will ignore Firefox.</para>

<para>To disable the currently installed Firefox, then install a new
Firefox while the old remains part of the profile:

<screen>
$ nix-env -q
firefox-2.0.0.9 <lineannotation>(the current one)</lineannotation>

$ nix-env --preserve-installed -i firefox-2.0.0.11
installing `firefox-2.0.0.11'
building path(s) `/nix/store/myy0y59q3ig70dgq37jqwg1j0rsapzsl-user-environment'
collision between `/nix/store/<replaceable>...</replaceable>-firefox-2.0.0.11/bin/firefox'
  and `/nix/store/<replaceable>...</replaceable>-firefox-2.0.0.9/bin/firefox'.
<lineannotation>(i.e., can&#x2019;t have two active at the same time)</lineannotation>

$ nix-env --set-flag active false firefox
setting flag on `firefox-2.0.0.9'

$ nix-env --preserve-installed -i firefox-2.0.0.11
installing `firefox-2.0.0.11'

$ nix-env -q
firefox-2.0.0.11 <lineannotation>(the enabled one)</lineannotation>
firefox-2.0.0.9 <lineannotation>(the disabled one)</lineannotation></screen>

</para>

<para>To make files from <literal>binutils</literal> take precedence
over files from <literal>gcc</literal>:

<screen>
$ nix-env --set-flag priority 5 binutils
$ nix-env --set-flag priority 10 gcc</screen>

</para>

</refsection>

</refsection>



<!--######################################################################-->

<refsection><title>Operation <option>--query</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-env</command>
  <group choice="req">
    <arg choice="plain"><option>--query</option></arg>
    <arg choice="plain"><option>-q</option></arg>
  </group>
  <group choice="opt">
    <arg choice="plain"><option>--installed</option></arg>
    <arg choice="plain"><option>--available</option></arg>
    <arg choice="plain"><option>-a</option></arg>
  </group>

  <sbr/>

  <arg>
    <group choice="req">
      <arg choice="plain"><option>--status</option></arg>
      <arg choice="plain"><option>-s</option></arg>
    </group>
  </arg>
  <arg>
    <group choice="req">
      <arg choice="plain"><option>--attr-path</option></arg>
      <arg choice="plain"><option>-P</option></arg>
    </group>
  </arg>
  <arg><option>--no-name</option></arg>
  <arg>
    <group choice="req">
      <arg choice="plain"><option>--compare-versions</option></arg>
      <arg choice="plain"><option>-c</option></arg>
    </group>
  </arg>
  <arg><option>--system</option></arg>
  <arg><option>--drv-path</option></arg>
  <arg><option>--out-path</option></arg>
  <arg><option>--description</option></arg>
  <arg><option>--meta</option></arg>

  <sbr/>

  <arg><option>--xml</option></arg>
  <arg><option>--json</option></arg>
  <arg>
    <group choice="req">
      <arg choice="plain"><option>--prebuilt-only</option></arg>
      <arg choice="plain"><option>-b</option></arg>
    </group>
  </arg>

  <arg>
    <group choice="req">
      <arg choice="plain"><option>--attr</option></arg>
      <arg choice="plain"><option>-A</option></arg>
    </group>
    <replaceable>attribute-path</replaceable>
  </arg>

  <sbr/>

  <arg choice="plain" rep="repeat"><replaceable>names</replaceable></arg>
</cmdsynopsis>

</refsection>


<refsection><title>Description</title>

<para>The query operation displays information about either the store
paths that are installed in the current generation of the active
profile (<option>--installed</option>), or the derivations that are
available for installation in the active Nix expression
(<option>--available</option>).  It only prints information about
derivations whose symbolic name matches one of
<replaceable>names</replaceable>.</para>

<para>The derivations are sorted by their <literal>name</literal>
attributes.</para>

</refsection>


<refsection><title>Source selection</title>

<para>The following flags specify the set of things on which the query
operates.</para>

<variablelist>

  <varlistentry><term><option>--installed</option></term>

    <listitem><para>The query operates on the store paths that are
    installed in the current generation of the active profile.  This
    is the default.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--available</option></term>
    <term><option>-a</option></term>

    <listitem><para>The query operates on the derivations that are
    available in the active Nix expression.</para></listitem>

  </varlistentry>

</variablelist>

</refsection>


<refsection><title>Queries</title>

<para>The following flags specify what information to display about
the selected derivations.  Multiple flags may be specified, in which
case the information is shown in the order given here.  Note that the
name of the derivation is shown unless <option>--no-name</option> is
specified.</para>

<!-- TODO: fix the terminology here; i.e., derivations, store paths,
user environment elements, etc. -->

<variablelist>

  <varlistentry><term><option>--xml</option></term>

    <listitem><para>Print the result in an XML representation suitable
    for automatic processing by other tools.  The root element is
    called <literal>items</literal>, which contains a
    <literal>item</literal> element for each available or installed
    derivation.  The fields discussed below are all stored in
    attributes of the <literal>item</literal>
    elements.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--json</option></term>

    <listitem><para>Print the result in a JSON representation suitable
    for automatic processing by other tools.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--prebuilt-only</option> / <option>-b</option></term>

    <listitem><para>Show only derivations for which a substitute is
    registered, i.e., there is a pre-built binary available that can
    be downloaded in lieu of building the derivation.  Thus, this
    shows all packages that probably can be installed
    quickly.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--status</option></term>
    <term><option>-s</option></term>

    <listitem><para>Print the <emphasis>status</emphasis> of the
    derivation.  The status consists of three characters.  The first
    is <literal>I</literal> or <literal>-</literal>, indicating
    whether the derivation is currently installed in the current
    generation of the active profile.  This is by definition the case
    for <option>--installed</option>, but not for
    <option>--available</option>.  The second is <literal>P</literal>
    or <literal>-</literal>, indicating whether the derivation is
    present on the system.  This indicates whether installation of an
    available derivation will require the derivation to be built.  The
    third is <literal>S</literal> or <literal>-</literal>, indicating
    whether a substitute is available for the
    derivation.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--attr-path</option></term>
    <term><option>-P</option></term>

    <listitem><para>Print the <emphasis>attribute path</emphasis> of
    the derivation, which can be used to unambiguously select it using
    the <link linkend="opt-attr"><option>--attr</option> option</link>
    available in commands that install derivations like
    <literal>nix-env --install</literal>. This option only works
    together with <option>--available</option></para></listitem>

  </varlistentry>

  <varlistentry><term><option>--no-name</option></term>

    <listitem><para>Suppress printing of the <literal>name</literal>
    attribute of each derivation.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--compare-versions</option> /
  <option>-c</option></term>

    <listitem><para>Compare installed versions to available versions,
    or vice versa (if <option>--available</option> is given).  This is
    useful for quickly seeing whether upgrades for installed
    packages are available in a Nix expression.  A column is added
    with the following meaning:

    <variablelist>

      <varlistentry><term><literal>&lt;</literal> <replaceable>version</replaceable></term>

        <listitem><para>A newer version of the package is available
        or installed.</para></listitem>

      </varlistentry>

      <varlistentry><term><literal>=</literal> <replaceable>version</replaceable></term>

        <listitem><para>At most the same version of the package is
        available or installed.</para></listitem>

      </varlistentry>

      <varlistentry><term><literal>&gt;</literal> <replaceable>version</replaceable></term>

        <listitem><para>Only older versions of the package are
        available or installed.</para></listitem>

      </varlistentry>

      <varlistentry><term><literal>- ?</literal></term>

        <listitem><para>No version of the package is available or
        installed.</para></listitem>

      </varlistentry>

    </variablelist>

    </para></listitem>

  </varlistentry>

  <varlistentry><term><option>--system</option></term>

    <listitem><para>Print the <literal>system</literal> attribute of
    the derivation.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--drv-path</option></term>

    <listitem><para>Print the path of the store
    derivation.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--out-path</option></term>

    <listitem><para>Print the output path of the
    derivation.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--description</option></term>

    <listitem><para>Print a short (one-line) description of the
    derivation, if available.  The description is taken from the
    <literal>meta.description</literal> attribute of the
    derivation.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--meta</option></term>

    <listitem><para>Print all of the meta-attributes of the
    derivation.  This option is only available with
    <option>--xml</option> or <option>--json</option>.</para></listitem>

  </varlistentry>

</variablelist>

</refsection>


<refsection><title>Examples</title>

<para>To show installed packages:

<screen>
$ nix-env -q
bison-1.875c
docbook-xml-4.2
firefox-1.0.4
MPlayer-1.0pre7
ORBit2-2.8.3
<replaceable>&#x2026;</replaceable>
</screen>

</para>

<para>To show available packages:

<screen>
$ nix-env -qa
firefox-1.0.7
GConf-2.4.0.1
MPlayer-1.0pre7
ORBit2-2.8.3
<replaceable>&#x2026;</replaceable>
</screen>

</para>

<para>To show the status of available packages:

<screen>
$ nix-env -qas
-P- firefox-1.0.7   <lineannotation>(not installed but present)</lineannotation>
--S GConf-2.4.0.1   <lineannotation>(not present, but there is a substitute for fast installation)</lineannotation>
--S MPlayer-1.0pre3 <lineannotation>(i.e., this is not the installed MPlayer, even though the version is the same!)</lineannotation>
IP- ORBit2-2.8.3    <lineannotation>(installed and by definition present)</lineannotation>
<replaceable>&#x2026;</replaceable>
</screen>

</para>

<para>To show available packages in the Nix expression <filename>foo.nix</filename>:

<screen>
$ nix-env -f ./foo.nix -qa
foo-1.2.3
</screen>

</para>

<para>To compare installed versions to what&#x2019;s available:

<screen>
$ nix-env -qc
<replaceable>...</replaceable>
acrobat-reader-7.0 - ?      <lineannotation>(package is not available at all)</lineannotation>
autoconf-2.59      = 2.59   <lineannotation>(same version)</lineannotation>
firefox-1.0.4      &lt; 1.0.7  <lineannotation>(a more recent version is available)</lineannotation>
<replaceable>...</replaceable>
</screen>

</para>

<para>To show all packages with &#x201C;<literal>zip</literal>&#x201D; in the name:

<screen>
$ nix-env -qa '.*zip.*'
bzip2-1.0.6
gzip-1.6
zip-3.0
<replaceable>&#x2026;</replaceable>
</screen>

</para>

<para>To show all packages with &#x201C;<literal>firefox</literal>&#x201D; or
&#x201C;<literal>chromium</literal>&#x201D; in the name:

<screen>
$ nix-env -qa '.*(firefox|chromium).*'
chromium-37.0.2062.94
chromium-beta-38.0.2125.24
firefox-32.0.3
firefox-with-plugins-13.0.1
<replaceable>&#x2026;</replaceable>
</screen>

</para>

<para>To show all packages in the latest revision of the Nixpkgs
repository:

<screen>
$ nix-env -f https://github.com/NixOS/nixpkgs/archive/master.tar.gz -qa
</screen>

</para>

</refsection>

</refsection>



<!--######################################################################-->

<refsection><title>Operation <option>--switch-profile</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-env</command>
  <group choice="req">
    <arg choice="plain"><option>--switch-profile</option></arg>
    <arg choice="plain"><option>-S</option></arg>
  </group>
  <arg choice="req"><replaceable>path</replaceable></arg>
</cmdsynopsis>

</refsection>


<refsection><title>Description</title>

<para>This operation makes <replaceable>path</replaceable> the current
profile for the user.  That is, the symlink
<filename>~/.nix-profile</filename> is made to point to
<replaceable>path</replaceable>.</para>

</refsection>

<refsection><title>Examples</title>

<screen>
$ nix-env -S ~/my-profile</screen>

</refsection>

</refsection>



<!--######################################################################-->

<refsection><title>Operation <option>--list-generations</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-env</command>
  <arg choice="plain"><option>--list-generations</option></arg>
</cmdsynopsis>

</refsection>


<refsection><title>Description</title>

<para>This operation print a list of all the currently existing
generations for the active profile.  These may be switched to using
the <option>--switch-generation</option> operation.  It also prints
the creation date of the generation, and indicates the current
generation.</para>

</refsection>


<refsection><title>Examples</title>

<screen>
$ nix-env --list-generations
  95   2004-02-06 11:48:24
  96   2004-02-06 11:49:01
  97   2004-02-06 16:22:45
  98   2004-02-06 16:24:33   (current)</screen>

</refsection>

</refsection>



<!--######################################################################-->

<refsection><title>Operation <option>--delete-generations</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-env</command>
  <arg choice="plain"><option>--delete-generations</option></arg>
  <arg choice="plain" rep="repeat"><replaceable>generations</replaceable></arg>
</cmdsynopsis>

</refsection>


<refsection><title>Description</title>

<para>This operation deletes the specified generations of the current
profile.  The generations can be a list of generation numbers, the
special value <literal>old</literal> to delete all non-current
generations,  a value such as <literal>30d</literal> to delete all
generations older than the specified number of days (except for the
generation that was active at that point in time), or a value such as
<literal>+5</literal> to keep the last <literal>5</literal> generations
ignoring any newer than current, e.g., if <literal>30</literal> is the current
generation <literal>+5</literal> will delete generation <literal>25</literal>
and all older generations.
Periodically deleting old generations is important to make garbage collection
effective.</para>

</refsection>

<refsection><title>Examples</title>

<screen>
$ nix-env --delete-generations 3 4 8

$ nix-env --delete-generations +5

$ nix-env --delete-generations 30d

$ nix-env -p other_profile --delete-generations old</screen>

</refsection>

</refsection>



<!--######################################################################-->

<refsection><title>Operation <option>--switch-generation</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-env</command>
  <group choice="req">
    <arg choice="plain"><option>--switch-generation</option></arg>
    <arg choice="plain"><option>-G</option></arg>
  </group>
  <arg choice="req"><replaceable>generation</replaceable></arg>
</cmdsynopsis>

</refsection>


<refsection><title>Description</title>

<para>This operation makes generation number
<replaceable>generation</replaceable> the current generation of the
active profile.  That is, if the
<filename><replaceable>profile</replaceable></filename> is the path to
the active profile, then the symlink
<filename><replaceable>profile</replaceable></filename> is made to
point to
<filename><replaceable>profile</replaceable>-<replaceable>generation</replaceable>-link</filename>,
which is in turn a symlink to the actual user environment in the Nix
store.</para>

<para>Switching will fail if the specified generation does not exist.</para>

</refsection>


<refsection><title>Examples</title>

<screen>
$ nix-env -G 42
switching from generation 50 to 42</screen>

</refsection>

</refsection>



<!--######################################################################-->

<refsection><title>Operation <option>--rollback</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-env</command>
  <arg choice="plain"><option>--rollback</option></arg>
</cmdsynopsis>

</refsection>

<refsection><title>Description</title>

<para>This operation switches to the &#x201C;previous&#x201D; generation of the
active profile, that is, the highest numbered generation lower than
the current generation, if it exists.  It is just a convenience
wrapper around <option>--list-generations</option> and
<option>--switch-generation</option>.</para>

</refsection>


<refsection><title>Examples</title>

<screen>
$ nix-env --rollback
switching from generation 92 to 91

$ nix-env --rollback
error: no generation older than the current (91) exists</screen>

</refsection>

</refsection>


<refsection condition="manpage"><title>Environment variables</title>

<variablelist>

  <varlistentry><term><envar>NIX_PROFILE</envar></term>

    <listitem><para>Location of the Nix profile.  Defaults to the
    target of the symlink <filename>~/.nix-profile</filename>, if it
    exists, or <filename>/nix/var/nix/profiles/default</filename>
    otherwise.</para></listitem>

  </varlistentry>

  <varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>IN_NIX_SHELL</envar></term>

  <listitem><para>Indicator that tells if the current environment was set up by
  <command>nix-shell</command>.  Since Nix 2.0 the values are
  <literal>"pure"</literal> and <literal>"impure"</literal></para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="env-NIX_PATH"><term><envar>NIX_PATH</envar></term>

  <listitem>

    <para>A colon-separated list of directories used to look up Nix
    expressions enclosed in angle brackets (i.e.,
    <literal>&lt;<replaceable>path</replaceable>&gt;</literal>).  For
    instance, the value

    <screen>
/home/eelco/Dev:/etc/nixos</screen>

    will cause Nix to look for paths relative to
    <filename>/home/eelco/Dev</filename> and
    <filename>/etc/nixos</filename>, in this order.  It is also
    possible to match paths against a prefix.  For example, the value

    <screen>
nixpkgs=/home/eelco/Dev/nixpkgs-branch:/etc/nixos</screen>

    will cause Nix to search for
    <literal>&lt;nixpkgs/<replaceable>path</replaceable>&gt;</literal> in
    <filename>/home/eelco/Dev/nixpkgs-branch/<replaceable>path</replaceable></filename>
    and
    <filename>/etc/nixos/nixpkgs/<replaceable>path</replaceable></filename>.</para>

    <para>If a path in the Nix search path starts with
    <literal>http://</literal> or <literal>https://</literal>, it is
    interpreted as the URL of a tarball that will be downloaded and
    unpacked to a temporary location. The tarball must consist of a
    single top-level directory. For example, setting
    <envar>NIX_PATH</envar> to

    <screen>
nixpkgs=https://github.com/NixOS/nixpkgs/archive/nixos-15.09.tar.gz</screen>

    tells Nix to download the latest revision in the Nixpkgs/NixOS
    15.09 channel.</para>

    <para>A following shorthand can be used to refer to the official channels:

    <screen>nixpkgs=channel:nixos-15.09</screen>
    </para>

    <para>The search path can be extended using the <option linkend="opt-I">-I</option> option, which takes precedence over
    <envar>NIX_PATH</envar>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_IGNORE_SYMLINK_STORE</envar></term>

  <listitem>

  <para>Normally, the Nix store directory (typically
  <filename>/nix/store</filename>) is not allowed to contain any
  symlink components.  This is to prevent &#x201C;impure&#x201D; builds.  Builders
  sometimes &#x201C;canonicalise&#x201D; paths by resolving all symlink components.
  Thus, builds on different machines (with
  <filename>/nix/store</filename> resolving to different locations)
  could yield different results.  This is generally not a problem,
  except when builds are deployed to machines where
  <filename>/nix/store</filename> resolves differently.  If you are
  sure that you&#x2019;re not going to do that, you can set
  <envar>NIX_IGNORE_SYMLINK_STORE</envar> to <envar>1</envar>.</para>

  <para>Note that if you&#x2019;re symlinking the Nix store so that you can
  put it on another file system than the root file system, on Linux
  you&#x2019;re better off using <literal>bind</literal> mount points, e.g.,

  <screen>
$ mkdir /nix
$ mount -o bind /mnt/otherdisk/nix /nix</screen>

  Consult the <citerefentry><refentrytitle>mount</refentrytitle>
  <manvolnum>8</manvolnum></citerefentry> manual page for details.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_STORE_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix store (default
  <filename><replaceable>prefix</replaceable>/store</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_DATA_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix static data
  directory (default
  <filename><replaceable>prefix</replaceable>/share</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_LOG_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix log directory
  (default <filename><replaceable>prefix</replaceable>/var/log/nix</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_STATE_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix state directory
  (default <filename><replaceable>prefix</replaceable>/var/nix</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_CONF_DIR</envar></term>

  <listitem><para>Overrides the location of the system Nix configuration
  directory (default
  <filename><replaceable>prefix</replaceable>/etc/nix</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_USER_CONF_FILES</envar></term>

  <listitem><para>Overrides the location of the user Nix configuration files
  to load from (defaults to the XDG spec locations). The variable is treated
  as a list separated by the <literal>:</literal> token.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>TMPDIR</envar></term>

  <listitem><para>Use the specified directory to store temporary
  files.  In particular, this includes temporary build directories;
  these can take up substantial amounts of disk space.  The default is
  <filename>/tmp</filename>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="envar-remote"><term><envar>NIX_REMOTE</envar></term>

  <listitem><para>This variable should be set to
  <literal>daemon</literal> if you want to use the Nix daemon to
  execute Nix operations. This is necessary in <link linkend="ssec-multi-user">multi-user Nix installations</link>.
  If the Nix daemon's Unix socket is at some non-standard path,
  this variable should be set to <literal>unix://path/to/socket</literal>.
  Otherwise, it should be left unset.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_SHOW_STATS</envar></term>

  <listitem><para>If set to <literal>1</literal>, Nix will print some
  evaluation statistics, such as the number of values
  allocated.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_COUNT_CALLS</envar></term>

  <listitem><para>If set to <literal>1</literal>, Nix will print how
  often functions were called during Nix expression evaluation.  This
  is useful for profiling your Nix expressions.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>GC_INITIAL_HEAP_SIZE</envar></term>

  <listitem><para>If Nix has been configured to use the Boehm garbage
  collector, this variable sets the initial size of the heap in bytes.
  It defaults to 384 MiB.  Setting it to a low value reduces memory
  consumption, but will increase runtime due to the overhead of
  garbage collection.</para></listitem>

</varlistentry>
</variablelist>

</refsection>


</refentry>
<refentry xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-nix-build">

<refmeta>
  <refentrytitle>nix-build</refentrytitle>
  <manvolnum>1</manvolnum>
  <refmiscinfo class="source">Nix</refmiscinfo>
  <refmiscinfo class="version">3.0</refmiscinfo>
</refmeta>

<refnamediv>
  <refname>nix-build</refname>
  <refpurpose>build a Nix expression</refpurpose>
</refnamediv>

<refsynopsisdiv>
  <cmdsynopsis>
    <command>nix-build</command>
    <arg xmlns="http://docbook.org/ns/docbook"><option>--help</option></arg><arg xmlns="http://docbook.org/ns/docbook"><option>--version</option></arg><arg xmlns="http://docbook.org/ns/docbook" rep="repeat">
  <group choice="req">
    <arg choice="plain"><option>--verbose</option></arg>
    <arg choice="plain"><option>-v</option></arg>
  </group>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <arg choice="plain"><option>--quiet</option></arg>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>--log-format</option>
  <replaceable>format</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <group choice="plain">
    <arg choice="plain"><option>--no-build-output</option></arg>
    <arg choice="plain"><option>-Q</option></arg>
  </group>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <group choice="req">
    <arg choice="plain"><option>--max-jobs</option></arg>
    <arg choice="plain"><option>-j</option></arg>
  </group>
  <replaceable>number</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>--cores</option>
  <replaceable>number</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>--max-silent-time</option>
  <replaceable>number</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>--timeout</option>
  <replaceable>number</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <group choice="plain">
    <arg choice="plain"><option>--keep-going</option></arg>
    <arg choice="plain"><option>-k</option></arg>
  </group>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <group choice="plain">
    <arg choice="plain"><option>--keep-failed</option></arg>
    <arg choice="plain"><option>-K</option></arg>
  </group>
</arg><arg xmlns="http://docbook.org/ns/docbook"><option>--fallback</option></arg><arg xmlns="http://docbook.org/ns/docbook"><option>--readonly-mode</option></arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>-I</option>
  <replaceable>path</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>--option</option>
  <replaceable>name</replaceable>
  <replaceable>value</replaceable>
</arg><sbr xmlns="http://docbook.org/ns/docbook"/>
    <arg><option>--arg</option> <replaceable>name</replaceable> <replaceable>value</replaceable></arg>
    <arg><option>--argstr</option> <replaceable>name</replaceable> <replaceable>value</replaceable></arg>
    <arg>
      <group choice="req">
        <arg choice="plain"><option>--attr</option></arg>
        <arg choice="plain"><option>-A</option></arg>
      </group>
      <replaceable>attrPath</replaceable>
    </arg>
    <arg><option>--no-out-link</option></arg>
    <arg><option>--dry-run</option></arg>
    <arg>
      <group choice="req">
        <arg choice="plain"><option>--out-link</option></arg>
        <arg choice="plain"><option>-o</option></arg>
      </group>
      <replaceable>outlink</replaceable>
    </arg>
    <arg choice="plain" rep="repeat"><replaceable>paths</replaceable></arg>
  </cmdsynopsis>
</refsynopsisdiv>

<refsection><title>Description</title>

<para>The <command>nix-build</command> command builds the derivations
described by the Nix expressions in <replaceable>paths</replaceable>.
If the build succeeds, it places a symlink to the result in the
current directory.  The symlink is called <filename>result</filename>.
If there are multiple Nix expressions, or the Nix expressions evaluate
to multiple derivations, multiple sequentially numbered symlinks are
created (<filename>result</filename>, <filename>result-2</filename>,
and so on).</para>

<para>If no <replaceable>paths</replaceable> are specified, then
<command>nix-build</command> will use <filename>default.nix</filename>
in the current directory, if it exists.</para>

<para>If an element of <replaceable>paths</replaceable> starts with
<literal>http://</literal> or <literal>https://</literal>, it is
interpreted as the URL of a tarball that will be downloaded and
unpacked to a temporary location. The tarball must include a single
top-level directory containing at least a file named
<filename>default.nix</filename>.</para>

<para><command>nix-build</command> is essentially a wrapper around
<link linkend="sec-nix-instantiate"><command>nix-instantiate</command></link>
(to translate a high-level Nix expression to a low-level store
derivation) and <link linkend="rsec-nix-store-realise"><command>nix-store
--realise</command></link> (to build the store derivation).</para>

<warning><para>The result of the build is automatically registered as
a root of the Nix garbage collector.  This root disappears
automatically when the <filename>result</filename> symlink is deleted
or renamed.  So don&#x2019;t rename the symlink.</para></warning>

</refsection>


<refsection><title>Options</title>

<para>All options not listed here are passed to <command>nix-store
--realise</command>, except for <option>--arg</option> and
<option>--attr</option> / <option>-A</option> which are passed to
<command>nix-instantiate</command>.  <phrase condition="manual">See
also <xref linkend="sec-common-options"/>.</phrase></para>

<variablelist>

  <varlistentry><term><option>--no-out-link</option></term>

    <listitem><para>Do not create a symlink to the output path.  Note
    that as a result the output does not become a root of the garbage
    collector, and so might be deleted by <command>nix-store
    --gc</command>.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--dry-run</option></term>
   <listitem><para>Show what store paths would be built or downloaded.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="opt-out-link"><term><option>--out-link</option> /
  <option>-o</option> <replaceable>outlink</replaceable></term>

    <listitem><para>Change the name of the symlink to the output path
    created from <filename>result</filename> to
    <replaceable>outlink</replaceable>.</para></listitem>

  </varlistentry>

</variablelist>

<para>The following common options are supported:</para>

<variablelist condition="manpage">
  <varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--help</option></term>

  <listitem><para>Prints out a summary of the command syntax and
  exits.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--version</option></term>

  <listitem><para>Prints out the Nix version number on standard output
  and exits.</para></listitem>
</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--verbose</option> / <option>-v</option></term>

  <listitem>

  <para>Increases the level of verbosity of diagnostic messages
  printed on standard error.  For each Nix operation, the information
  printed on standard output is well-defined; any diagnostic
  information is printed on standard error, never on standard
  output.</para>

  <para>This option may be specified repeatedly.  Currently, the
  following verbosity levels exist:</para>

  <variablelist>

    <varlistentry><term>0</term>
    <listitem><para>&#x201C;Errors only&#x201D;: only print messages
    explaining why the Nix invocation failed.</para></listitem>
    </varlistentry>

    <varlistentry><term>1</term>
    <listitem><para>&#x201C;Informational&#x201D;: print
    <emphasis>useful</emphasis> messages about what Nix is doing.
    This is the default.</para></listitem>
    </varlistentry>

    <varlistentry><term>2</term>
    <listitem><para>&#x201C;Talkative&#x201D;: print more informational
    messages.</para></listitem>
    </varlistentry>

    <varlistentry><term>3</term>
    <listitem><para>&#x201C;Chatty&#x201D;: print even more
    informational messages.</para></listitem>
    </varlistentry>

    <varlistentry><term>4</term>
    <listitem><para>&#x201C;Debug&#x201D;: print debug
    information.</para></listitem>
    </varlistentry>

    <varlistentry><term>5</term>
    <listitem><para>&#x201C;Vomit&#x201D;: print vast amounts of debug
    information.</para></listitem>
    </varlistentry>

  </variablelist>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--quiet</option></term>

  <listitem>

  <para>Decreases the level of verbosity of diagnostic messages
  printed on standard error.  This is the inverse option to
  <option>-v</option> / <option>--verbose</option>.
  </para>

  <para>This option may be specified repeatedly.  See the previous
  verbosity levels list.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-log-format"><term><option>--log-format</option> <replaceable>format</replaceable></term>

  <listitem>

  <para>This option can be used to change the output of the log format, with
  <replaceable>format</replaceable> being one of:</para>

  <variablelist>

    <varlistentry><term>raw</term>
    <listitem><para>This is the raw format, as outputted by nix-build.</para></listitem>
    </varlistentry>

    <varlistentry><term>internal-json</term>
    <listitem><para>Outputs the logs in a structured manner. NOTE: the json schema is not guarantees to be stable between releases.</para></listitem>
    </varlistentry>

    <varlistentry><term>bar</term>
    <listitem><para>Only display a progress bar during the builds.</para></listitem>
    </varlistentry>

    <varlistentry><term>bar-with-logs</term>
    <listitem><para>Display the raw logs, with the progress bar at the bottom.</para></listitem>
    </varlistentry>

  </variablelist>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--no-build-output</option> / <option>-Q</option></term>

  <listitem><para>By default, output written by builders to standard
  output and standard error is echoed to the Nix command's standard
  error.  This option suppresses this behaviour.  Note that the
  builder's standard output and error are always written to a log file
  in
  <filename><replaceable>prefix</replaceable>/nix/var/log/nix</filename>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-max-jobs"><term><option>--max-jobs</option> / <option>-j</option>
<replaceable>number</replaceable></term>

  <listitem>

  <para>Sets the maximum number of build jobs that Nix will
  perform in parallel to the specified number.  Specify
  <literal>auto</literal> to use the number of CPUs in the system.
  The default is specified by the <link linkend="conf-max-jobs"><literal>max-jobs</literal></link>
  configuration setting, which itself defaults to
  <literal>1</literal>.  A higher value is useful on SMP systems or to
  exploit I/O latency.</para>

  <para> Setting it to <literal>0</literal> disallows building on the local
  machine, which is useful when you want builds to happen only on remote
  builders.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-cores"><term><option>--cores</option></term>

  <listitem><para>Sets the value of the <envar>NIX_BUILD_CORES</envar>
  environment variable in the invocation of builders.  Builders can
  use this variable at their discretion to control the maximum amount
  of parallelism.  For instance, in Nixpkgs, if the derivation
  attribute <varname>enableParallelBuilding</varname> is set to
  <literal>true</literal>, the builder passes the
  <option>-j<replaceable>N</replaceable></option> flag to GNU Make.
  It defaults to the value of the <link linkend="conf-cores"><literal>cores</literal></link>
  configuration setting, if set, or <literal>1</literal> otherwise.
  The value <literal>0</literal> means that the builder should use all
  available CPU cores in the system.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-max-silent-time"><term><option>--max-silent-time</option></term>

  <listitem><para>Sets the maximum number of seconds that a builder
  can go without producing any data on standard output or standard
  error.  The default is specified by the <link linkend="conf-max-silent-time"><literal>max-silent-time</literal></link>
  configuration setting.  <literal>0</literal> means no
  time-out.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-timeout"><term><option>--timeout</option></term>

  <listitem><para>Sets the maximum number of seconds that a builder
  can run.  The default is specified by the <link linkend="conf-timeout"><literal>timeout</literal></link>
  configuration setting.  <literal>0</literal> means no
  timeout.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--keep-going</option> / <option>-k</option></term>

  <listitem><para>Keep going in case of failed builds, to the
  greatest extent possible.  That is, if building an input of some
  derivation fails, Nix will still build the other inputs, but not the
  derivation itself.  Without this option, Nix stops if any build
  fails (except for builds of substitutes), possibly killing builds in
  progress (in case of parallel or distributed builds).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--keep-failed</option> / <option>-K</option></term>

  <listitem><para>Specifies that in case of a build failure, the
  temporary directory (usually in <filename>/tmp</filename>) in which
  the build takes place should not be deleted.  The path of the build
  directory is printed as an informational message.
    </para>
  </listitem>
</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--fallback</option></term>

  <listitem>

  <para>Whenever Nix attempts to build a derivation for which
  substitutes are known for each output path, but realising the output
  paths through the substitutes fails, fall back on building the
  derivation.</para>

  <para>The most common scenario in which this is useful is when we
  have registered substitutes in order to perform binary distribution
  from, say, a network repository.  If the repository is down, the
  realisation of the derivation will fail.  When this option is
  specified, Nix will build the derivation instead.  Thus,
  installation from binaries falls back on installation from source.
  This option is not the default since it is generally not desirable
  for a transient failure in obtaining the substitutes to lead to a
  full build from source (with the related consumption of
  resources).</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--no-build-hook</option></term>

  <listitem>

  <para>Disables the build hook mechanism.  This allows to ignore remote
  builders if they are setup on the machine.</para>

  <para>It's useful in cases where the bandwidth between the client and the
  remote builder is too low.  In that case it can take more time to upload the
  sources to the remote builder and fetch back the result than to do the
  computation locally.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--readonly-mode</option></term>

  <listitem><para>When this option is used, no attempt is made to open
  the Nix database.  Most Nix operations do need database access, so
  those operations will fail.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--arg</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>This option is accepted by
  <command>nix-env</command>, <command>nix-instantiate</command>,
  <command>nix-shell</command> and <command>nix-build</command>.
  When evaluating Nix expressions, the expression evaluator will
  automatically try to call functions that
  it encounters.  It can automatically call functions for which every
  argument has a <link linkend="ss-functions">default value</link>
  (e.g., <literal>{ <replaceable>argName</replaceable> ?
  <replaceable>defaultValue</replaceable> }:
  <replaceable>...</replaceable></literal>).  With
  <option>--arg</option>, you can also call functions that have
  arguments without a default value (or override a default value).
  That is, if the evaluator encounters a function with an argument
  named <replaceable>name</replaceable>, it will call it with value
  <replaceable>value</replaceable>.</para>

  <para>For instance, the top-level <literal>default.nix</literal> in
  Nixpkgs is actually a function:

<programlisting>
{ # The system (e.g., `i686-linux') for which to build the packages.
  system ? builtins.currentSystem
  <replaceable>...</replaceable>
}: <replaceable>...</replaceable></programlisting>

  So if you call this Nix expression (e.g., when you do
  <literal>nix-env -i <replaceable>pkgname</replaceable></literal>),
  the function will be called automatically using the value <link linkend="builtin-currentSystem"><literal>builtins.currentSystem</literal></link>
  for the <literal>system</literal> argument.  You can override this
  using <option>--arg</option>, e.g., <literal>nix-env -i
  <replaceable>pkgname</replaceable> --arg system
  \"i686-freebsd\"</literal>.  (Note that since the argument is a Nix
  string literal, you have to escape the quotes.)</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--argstr</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>This option is like <option>--arg</option>, only the
  value is not a Nix expression but a string.  So instead of
  <literal>--arg system \"i686-linux\"</literal> (the outer quotes are
  to keep the shell happy) you can say <literal>--argstr system
  i686-linux</literal>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-attr"><term><option>--attr</option> / <option>-A</option>
<replaceable>attrPath</replaceable></term>

  <listitem><para>Select an attribute from the top-level Nix
  expression being evaluated.  (<command>nix-env</command>,
  <command>nix-instantiate</command>, <command>nix-build</command> and
  <command>nix-shell</command> only.)  The <emphasis>attribute
  path</emphasis> <replaceable>attrPath</replaceable> is a sequence of
  attribute names separated by dots.  For instance, given a top-level
  Nix expression <replaceable>e</replaceable>, the attribute path
  <literal>xorg.xorgserver</literal> would cause the expression
  <literal><replaceable>e</replaceable>.xorg.xorgserver</literal> to
  be used.  See <link linkend="refsec-nix-env-install-examples"><command>nix-env
  --install</command></link> for some concrete examples.</para>

  <para>In addition to attribute names, you can also specify array
  indices.  For instance, the attribute path
  <literal>foo.3.bar</literal> selects the <literal>bar</literal>
  attribute of the fourth element of the array in the
  <literal>foo</literal> attribute of the top-level
  expression.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--expr</option> / <option>-E</option></term>

  <listitem><para>Interpret the command line arguments as a list of
  Nix expressions to be parsed and evaluated, rather than as a list
  of file names of Nix expressions.
  (<command>nix-instantiate</command>, <command>nix-build</command>
  and <command>nix-shell</command> only.)</para>

  <para>For <command>nix-shell</command>, this option is commonly used
  to give you a shell in which you can build the packages returned
  by the expression. If you want to get a shell which contain the
  <emphasis>built</emphasis> packages ready for use, give your
  expression to the <command>nix-shell -p</command> convenience flag
  instead.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-I"><term><option>-I</option> <replaceable>path</replaceable></term>

  <listitem><para>Add a path to the Nix expression search path.  This
  option may be given multiple times.  See the <envar linkend="env-NIX_PATH">NIX_PATH</envar> environment variable for
  information on the semantics of the Nix search path.  Paths added
  through <option>-I</option> take precedence over
  <envar>NIX_PATH</envar>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--option</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>Set the Nix configuration option
  <replaceable>name</replaceable> to <replaceable>value</replaceable>.
  This overrides settings in the Nix configuration file (see
  <citerefentry><refentrytitle>nix.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--repair</option></term>

  <listitem><para>Fix corrupted or missing store paths by
  redownloading or rebuilding them.  Note that this is slow because it
  requires computing a cryptographic hash of the contents of every
  path in the closure of the build.  Also note the warning under
  <command>nix-store --repair-path</command>.</para></listitem>

</varlistentry>
</variablelist>

</refsection>


<refsection><title>Examples</title>

<screen>
$ nix-build '&lt;nixpkgs&gt;' -A firefox
store derivation is /nix/store/qybprl8sz2lc...-firefox-1.5.0.7.drv
/nix/store/d18hyl92g30l...-firefox-1.5.0.7

$ ls -l result
lrwxrwxrwx  <replaceable>...</replaceable>  result -&gt; /nix/store/d18hyl92g30l...-firefox-1.5.0.7

$ ls ./result/bin/
firefox  firefox-config</screen>

<para>If a derivation has multiple outputs,
<command>nix-build</command> will build the default (first) output.
You can also build all outputs:
<screen>
$ nix-build '&lt;nixpkgs&gt;' -A openssl.all
</screen>
This will create a symlink for each output named
<filename>result-<replaceable>outputname</replaceable></filename>.
The suffix is omitted if the output name is <literal>out</literal>.
So if <literal>openssl</literal> has outputs <literal>out</literal>,
<literal>bin</literal> and <literal>man</literal>,
<command>nix-build</command> will create symlinks
<literal>result</literal>, <literal>result-bin</literal> and
<literal>result-man</literal>.  It&#x2019;s also possible to build a specific
output:
<screen>
$ nix-build '&lt;nixpkgs&gt;' -A openssl.man
</screen>
This will create a symlink <literal>result-man</literal>.</para>

<para>Build a Nix expression given on the command line:

<screen>
$ nix-build -E 'with import &lt;nixpkgs&gt; { }; runCommand "foo" { } "echo bar &gt; $out"'
$ cat ./result
bar
</screen>

</para>

<para>Build the GNU Hello package from the latest revision of the
master branch of Nixpkgs:

<screen>
$ nix-build https://github.com/NixOS/nixpkgs/archive/master.tar.gz -A hello
</screen>

</para>

</refsection>


<refsection condition="manpage"><title>Environment variables</title>

<variablelist>
  <varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>IN_NIX_SHELL</envar></term>

  <listitem><para>Indicator that tells if the current environment was set up by
  <command>nix-shell</command>.  Since Nix 2.0 the values are
  <literal>"pure"</literal> and <literal>"impure"</literal></para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="env-NIX_PATH"><term><envar>NIX_PATH</envar></term>

  <listitem>

    <para>A colon-separated list of directories used to look up Nix
    expressions enclosed in angle brackets (i.e.,
    <literal>&lt;<replaceable>path</replaceable>&gt;</literal>).  For
    instance, the value

    <screen>
/home/eelco/Dev:/etc/nixos</screen>

    will cause Nix to look for paths relative to
    <filename>/home/eelco/Dev</filename> and
    <filename>/etc/nixos</filename>, in this order.  It is also
    possible to match paths against a prefix.  For example, the value

    <screen>
nixpkgs=/home/eelco/Dev/nixpkgs-branch:/etc/nixos</screen>

    will cause Nix to search for
    <literal>&lt;nixpkgs/<replaceable>path</replaceable>&gt;</literal> in
    <filename>/home/eelco/Dev/nixpkgs-branch/<replaceable>path</replaceable></filename>
    and
    <filename>/etc/nixos/nixpkgs/<replaceable>path</replaceable></filename>.</para>

    <para>If a path in the Nix search path starts with
    <literal>http://</literal> or <literal>https://</literal>, it is
    interpreted as the URL of a tarball that will be downloaded and
    unpacked to a temporary location. The tarball must consist of a
    single top-level directory. For example, setting
    <envar>NIX_PATH</envar> to

    <screen>
nixpkgs=https://github.com/NixOS/nixpkgs/archive/nixos-15.09.tar.gz</screen>

    tells Nix to download the latest revision in the Nixpkgs/NixOS
    15.09 channel.</para>

    <para>A following shorthand can be used to refer to the official channels:

    <screen>nixpkgs=channel:nixos-15.09</screen>
    </para>

    <para>The search path can be extended using the <option linkend="opt-I">-I</option> option, which takes precedence over
    <envar>NIX_PATH</envar>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_IGNORE_SYMLINK_STORE</envar></term>

  <listitem>

  <para>Normally, the Nix store directory (typically
  <filename>/nix/store</filename>) is not allowed to contain any
  symlink components.  This is to prevent &#x201C;impure&#x201D; builds.  Builders
  sometimes &#x201C;canonicalise&#x201D; paths by resolving all symlink components.
  Thus, builds on different machines (with
  <filename>/nix/store</filename> resolving to different locations)
  could yield different results.  This is generally not a problem,
  except when builds are deployed to machines where
  <filename>/nix/store</filename> resolves differently.  If you are
  sure that you&#x2019;re not going to do that, you can set
  <envar>NIX_IGNORE_SYMLINK_STORE</envar> to <envar>1</envar>.</para>

  <para>Note that if you&#x2019;re symlinking the Nix store so that you can
  put it on another file system than the root file system, on Linux
  you&#x2019;re better off using <literal>bind</literal> mount points, e.g.,

  <screen>
$ mkdir /nix
$ mount -o bind /mnt/otherdisk/nix /nix</screen>

  Consult the <citerefentry><refentrytitle>mount</refentrytitle>
  <manvolnum>8</manvolnum></citerefentry> manual page for details.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_STORE_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix store (default
  <filename><replaceable>prefix</replaceable>/store</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_DATA_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix static data
  directory (default
  <filename><replaceable>prefix</replaceable>/share</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_LOG_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix log directory
  (default <filename><replaceable>prefix</replaceable>/var/log/nix</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_STATE_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix state directory
  (default <filename><replaceable>prefix</replaceable>/var/nix</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_CONF_DIR</envar></term>

  <listitem><para>Overrides the location of the system Nix configuration
  directory (default
  <filename><replaceable>prefix</replaceable>/etc/nix</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_USER_CONF_FILES</envar></term>

  <listitem><para>Overrides the location of the user Nix configuration files
  to load from (defaults to the XDG spec locations). The variable is treated
  as a list separated by the <literal>:</literal> token.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>TMPDIR</envar></term>

  <listitem><para>Use the specified directory to store temporary
  files.  In particular, this includes temporary build directories;
  these can take up substantial amounts of disk space.  The default is
  <filename>/tmp</filename>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="envar-remote"><term><envar>NIX_REMOTE</envar></term>

  <listitem><para>This variable should be set to
  <literal>daemon</literal> if you want to use the Nix daemon to
  execute Nix operations. This is necessary in <link linkend="ssec-multi-user">multi-user Nix installations</link>.
  If the Nix daemon's Unix socket is at some non-standard path,
  this variable should be set to <literal>unix://path/to/socket</literal>.
  Otherwise, it should be left unset.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_SHOW_STATS</envar></term>

  <listitem><para>If set to <literal>1</literal>, Nix will print some
  evaluation statistics, such as the number of values
  allocated.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_COUNT_CALLS</envar></term>

  <listitem><para>If set to <literal>1</literal>, Nix will print how
  often functions were called during Nix expression evaluation.  This
  is useful for profiling your Nix expressions.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>GC_INITIAL_HEAP_SIZE</envar></term>

  <listitem><para>If Nix has been configured to use the Boehm garbage
  collector, this variable sets the initial size of the heap in bytes.
  It defaults to 384 MiB.  Setting it to a low value reduces memory
  consumption, but will increase runtime due to the overhead of
  garbage collection.</para></listitem>

</varlistentry>
</variablelist>

</refsection>


</refentry>
<refentry xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-nix-shell">

<refmeta>
  <refentrytitle>nix-shell</refentrytitle>
  <manvolnum>1</manvolnum>
  <refmiscinfo class="source">Nix</refmiscinfo>
  <refmiscinfo class="version">3.0</refmiscinfo>
</refmeta>

<refnamediv>
  <refname>nix-shell</refname>
  <refpurpose>start an interactive shell based on a Nix expression</refpurpose>
</refnamediv>

<refsynopsisdiv>
  <cmdsynopsis>
    <command>nix-shell</command>
    <arg><option>--arg</option> <replaceable>name</replaceable> <replaceable>value</replaceable></arg>
    <arg><option>--argstr</option> <replaceable>name</replaceable> <replaceable>value</replaceable></arg>
    <arg>
      <group choice="req">
        <arg choice="plain"><option>--attr</option></arg>
        <arg choice="plain"><option>-A</option></arg>
      </group>
      <replaceable>attrPath</replaceable>
    </arg>
    <arg><option>--command</option> <replaceable>cmd</replaceable></arg>
    <arg><option>--run</option> <replaceable>cmd</replaceable></arg>
    <arg><option>--exclude</option> <replaceable>regexp</replaceable></arg>
    <arg><option>--pure</option></arg>
    <arg><option>--keep</option> <replaceable>name</replaceable></arg>
    <group choice="req">
      <arg choice="plain">
        <group choice="req">
          <arg choice="plain"><option>--packages</option></arg>
          <arg choice="plain"><option>-p</option></arg>
        </group>
        <arg choice="plain" rep="repeat">
          <group choice="req">
            <arg choice="plain"><replaceable>packages</replaceable></arg>
            <arg choice="plain"><replaceable>expressions</replaceable></arg>
          </group>
        </arg>
      </arg>
      <arg><replaceable>path</replaceable></arg>
    </group>
  </cmdsynopsis>
</refsynopsisdiv>

<refsection><title>Description</title>

<para>The command <command>nix-shell</command> will build the
dependencies of the specified derivation, but not the derivation
itself.  It will then start an interactive shell in which all
environment variables defined by the derivation
<replaceable>path</replaceable> have been set to their corresponding
values, and the script <literal>$stdenv/setup</literal> has been
sourced.  This is useful for reproducing the environment of a
derivation for development.</para>

<para>If <replaceable>path</replaceable> is not given,
<command>nix-shell</command> defaults to
<filename>shell.nix</filename> if it exists, and
<filename>default.nix</filename> otherwise.</para>

<para>If <replaceable>path</replaceable> starts with
<literal>http://</literal> or <literal>https://</literal>, it is
interpreted as the URL of a tarball that will be downloaded and
unpacked to a temporary location. The tarball must include a single
top-level directory containing at least a file named
<filename>default.nix</filename>.</para>

<para>If the derivation defines the variable
<varname>shellHook</varname>, it will be evaluated after
<literal>$stdenv/setup</literal> has been sourced.  Since this hook is
not executed by regular Nix builds, it allows you to perform
initialisation specific to <command>nix-shell</command>.  For example,
the derivation attribute

<programlisting>
shellHook =
  ''
    echo "Hello shell"
  '';
</programlisting>

will cause <command>nix-shell</command> to print <literal>Hello shell</literal>.</para>

</refsection>


<refsection><title>Options</title>

<para>All options not listed here are passed to <command>nix-store
--realise</command>, except for <option>--arg</option> and
<option>--attr</option> / <option>-A</option> which are passed to
<command>nix-instantiate</command>.  <phrase condition="manual">See
also <xref linkend="sec-common-options"/>.</phrase></para>

<variablelist>

  <varlistentry><term><option>--command</option> <replaceable>cmd</replaceable></term>

    <listitem><para>In the environment of the derivation, run the
    shell command <replaceable>cmd</replaceable>. This command is
    executed in an interactive shell. (Use <option>--run</option> to
    use a non-interactive shell instead.) However, a call to
    <literal>exit</literal> is implicitly added to the command, so the
    shell will exit after running the command. To prevent this, add
    <literal>return</literal> at the end; e.g. <literal>--command
    "echo Hello; return"</literal> will print <literal>Hello</literal>
    and then drop you into the interactive shell. This can be useful
    for doing any additional initialisation.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--run</option> <replaceable>cmd</replaceable></term>

    <listitem><para>Like <option>--command</option>, but executes the
    command in a non-interactive shell. This means (among other
    things) that if you hit Ctrl-C while the command is running, the
    shell exits.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--exclude</option> <replaceable>regexp</replaceable></term>

    <listitem><para>Do not build any dependencies whose store path
    matches the regular expression <replaceable>regexp</replaceable>.
    This option may be specified multiple times.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--pure</option></term>

    <listitem><para>If this flag is specified, the environment is
    almost entirely cleared before the interactive shell is started,
    so you get an environment that more closely corresponds to the
    &#x201C;real&#x201D; Nix build.  A few variables, in particular
    <envar>HOME</envar>, <envar>USER</envar> and
    <envar>DISPLAY</envar>, are retained.  Note that
    <filename>~/.bashrc</filename> and (depending on your Bash
    installation) <filename>/etc/bashrc</filename> are still sourced,
    so any variables set there will affect the interactive
    shell.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--packages</option> / <option>-p</option> <replaceable>packages</replaceable>&#x2026;</term>

    <listitem><para>Set up an environment in which the specified
    packages are present.  The command line arguments are interpreted
    as attribute names inside the Nix Packages collection.  Thus,
    <literal>nix-shell -p libjpeg openjdk</literal> will start a shell
    in which the packages denoted by the attribute names
    <varname>libjpeg</varname> and <varname>openjdk</varname> are
    present.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>-i</option> <replaceable>interpreter</replaceable></term>

    <listitem><para>The chained script interpreter to be invoked by
    <command>nix-shell</command>. Only applicable in
    <literal>#!</literal>-scripts (described <link linkend="ssec-nix-shell-shebang">below</link>).</para>

    </listitem></varlistentry>

  <varlistentry><term><option>--keep</option> <replaceable>name</replaceable></term>

    <listitem><para>When a <option>--pure</option> shell is started,
    keep the listed environment variables.</para></listitem>

  </varlistentry>

</variablelist>

<para>The following common options are supported:</para>

<variablelist condition="manpage">
  <varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--help</option></term>

  <listitem><para>Prints out a summary of the command syntax and
  exits.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--version</option></term>

  <listitem><para>Prints out the Nix version number on standard output
  and exits.</para></listitem>
</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--verbose</option> / <option>-v</option></term>

  <listitem>

  <para>Increases the level of verbosity of diagnostic messages
  printed on standard error.  For each Nix operation, the information
  printed on standard output is well-defined; any diagnostic
  information is printed on standard error, never on standard
  output.</para>

  <para>This option may be specified repeatedly.  Currently, the
  following verbosity levels exist:</para>

  <variablelist>

    <varlistentry><term>0</term>
    <listitem><para>&#x201C;Errors only&#x201D;: only print messages
    explaining why the Nix invocation failed.</para></listitem>
    </varlistentry>

    <varlistentry><term>1</term>
    <listitem><para>&#x201C;Informational&#x201D;: print
    <emphasis>useful</emphasis> messages about what Nix is doing.
    This is the default.</para></listitem>
    </varlistentry>

    <varlistentry><term>2</term>
    <listitem><para>&#x201C;Talkative&#x201D;: print more informational
    messages.</para></listitem>
    </varlistentry>

    <varlistentry><term>3</term>
    <listitem><para>&#x201C;Chatty&#x201D;: print even more
    informational messages.</para></listitem>
    </varlistentry>

    <varlistentry><term>4</term>
    <listitem><para>&#x201C;Debug&#x201D;: print debug
    information.</para></listitem>
    </varlistentry>

    <varlistentry><term>5</term>
    <listitem><para>&#x201C;Vomit&#x201D;: print vast amounts of debug
    information.</para></listitem>
    </varlistentry>

  </variablelist>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--quiet</option></term>

  <listitem>

  <para>Decreases the level of verbosity of diagnostic messages
  printed on standard error.  This is the inverse option to
  <option>-v</option> / <option>--verbose</option>.
  </para>

  <para>This option may be specified repeatedly.  See the previous
  verbosity levels list.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-log-format"><term><option>--log-format</option> <replaceable>format</replaceable></term>

  <listitem>

  <para>This option can be used to change the output of the log format, with
  <replaceable>format</replaceable> being one of:</para>

  <variablelist>

    <varlistentry><term>raw</term>
    <listitem><para>This is the raw format, as outputted by nix-build.</para></listitem>
    </varlistentry>

    <varlistentry><term>internal-json</term>
    <listitem><para>Outputs the logs in a structured manner. NOTE: the json schema is not guarantees to be stable between releases.</para></listitem>
    </varlistentry>

    <varlistentry><term>bar</term>
    <listitem><para>Only display a progress bar during the builds.</para></listitem>
    </varlistentry>

    <varlistentry><term>bar-with-logs</term>
    <listitem><para>Display the raw logs, with the progress bar at the bottom.</para></listitem>
    </varlistentry>

  </variablelist>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--no-build-output</option> / <option>-Q</option></term>

  <listitem><para>By default, output written by builders to standard
  output and standard error is echoed to the Nix command's standard
  error.  This option suppresses this behaviour.  Note that the
  builder's standard output and error are always written to a log file
  in
  <filename><replaceable>prefix</replaceable>/nix/var/log/nix</filename>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-max-jobs"><term><option>--max-jobs</option> / <option>-j</option>
<replaceable>number</replaceable></term>

  <listitem>

  <para>Sets the maximum number of build jobs that Nix will
  perform in parallel to the specified number.  Specify
  <literal>auto</literal> to use the number of CPUs in the system.
  The default is specified by the <link linkend="conf-max-jobs"><literal>max-jobs</literal></link>
  configuration setting, which itself defaults to
  <literal>1</literal>.  A higher value is useful on SMP systems or to
  exploit I/O latency.</para>

  <para> Setting it to <literal>0</literal> disallows building on the local
  machine, which is useful when you want builds to happen only on remote
  builders.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-cores"><term><option>--cores</option></term>

  <listitem><para>Sets the value of the <envar>NIX_BUILD_CORES</envar>
  environment variable in the invocation of builders.  Builders can
  use this variable at their discretion to control the maximum amount
  of parallelism.  For instance, in Nixpkgs, if the derivation
  attribute <varname>enableParallelBuilding</varname> is set to
  <literal>true</literal>, the builder passes the
  <option>-j<replaceable>N</replaceable></option> flag to GNU Make.
  It defaults to the value of the <link linkend="conf-cores"><literal>cores</literal></link>
  configuration setting, if set, or <literal>1</literal> otherwise.
  The value <literal>0</literal> means that the builder should use all
  available CPU cores in the system.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-max-silent-time"><term><option>--max-silent-time</option></term>

  <listitem><para>Sets the maximum number of seconds that a builder
  can go without producing any data on standard output or standard
  error.  The default is specified by the <link linkend="conf-max-silent-time"><literal>max-silent-time</literal></link>
  configuration setting.  <literal>0</literal> means no
  time-out.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-timeout"><term><option>--timeout</option></term>

  <listitem><para>Sets the maximum number of seconds that a builder
  can run.  The default is specified by the <link linkend="conf-timeout"><literal>timeout</literal></link>
  configuration setting.  <literal>0</literal> means no
  timeout.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--keep-going</option> / <option>-k</option></term>

  <listitem><para>Keep going in case of failed builds, to the
  greatest extent possible.  That is, if building an input of some
  derivation fails, Nix will still build the other inputs, but not the
  derivation itself.  Without this option, Nix stops if any build
  fails (except for builds of substitutes), possibly killing builds in
  progress (in case of parallel or distributed builds).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--keep-failed</option> / <option>-K</option></term>

  <listitem><para>Specifies that in case of a build failure, the
  temporary directory (usually in <filename>/tmp</filename>) in which
  the build takes place should not be deleted.  The path of the build
  directory is printed as an informational message.
    </para>
  </listitem>
</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--fallback</option></term>

  <listitem>

  <para>Whenever Nix attempts to build a derivation for which
  substitutes are known for each output path, but realising the output
  paths through the substitutes fails, fall back on building the
  derivation.</para>

  <para>The most common scenario in which this is useful is when we
  have registered substitutes in order to perform binary distribution
  from, say, a network repository.  If the repository is down, the
  realisation of the derivation will fail.  When this option is
  specified, Nix will build the derivation instead.  Thus,
  installation from binaries falls back on installation from source.
  This option is not the default since it is generally not desirable
  for a transient failure in obtaining the substitutes to lead to a
  full build from source (with the related consumption of
  resources).</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--no-build-hook</option></term>

  <listitem>

  <para>Disables the build hook mechanism.  This allows to ignore remote
  builders if they are setup on the machine.</para>

  <para>It's useful in cases where the bandwidth between the client and the
  remote builder is too low.  In that case it can take more time to upload the
  sources to the remote builder and fetch back the result than to do the
  computation locally.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--readonly-mode</option></term>

  <listitem><para>When this option is used, no attempt is made to open
  the Nix database.  Most Nix operations do need database access, so
  those operations will fail.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--arg</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>This option is accepted by
  <command>nix-env</command>, <command>nix-instantiate</command>,
  <command>nix-shell</command> and <command>nix-build</command>.
  When evaluating Nix expressions, the expression evaluator will
  automatically try to call functions that
  it encounters.  It can automatically call functions for which every
  argument has a <link linkend="ss-functions">default value</link>
  (e.g., <literal>{ <replaceable>argName</replaceable> ?
  <replaceable>defaultValue</replaceable> }:
  <replaceable>...</replaceable></literal>).  With
  <option>--arg</option>, you can also call functions that have
  arguments without a default value (or override a default value).
  That is, if the evaluator encounters a function with an argument
  named <replaceable>name</replaceable>, it will call it with value
  <replaceable>value</replaceable>.</para>

  <para>For instance, the top-level <literal>default.nix</literal> in
  Nixpkgs is actually a function:

<programlisting>
{ # The system (e.g., `i686-linux') for which to build the packages.
  system ? builtins.currentSystem
  <replaceable>...</replaceable>
}: <replaceable>...</replaceable></programlisting>

  So if you call this Nix expression (e.g., when you do
  <literal>nix-env -i <replaceable>pkgname</replaceable></literal>),
  the function will be called automatically using the value <link linkend="builtin-currentSystem"><literal>builtins.currentSystem</literal></link>
  for the <literal>system</literal> argument.  You can override this
  using <option>--arg</option>, e.g., <literal>nix-env -i
  <replaceable>pkgname</replaceable> --arg system
  \"i686-freebsd\"</literal>.  (Note that since the argument is a Nix
  string literal, you have to escape the quotes.)</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--argstr</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>This option is like <option>--arg</option>, only the
  value is not a Nix expression but a string.  So instead of
  <literal>--arg system \"i686-linux\"</literal> (the outer quotes are
  to keep the shell happy) you can say <literal>--argstr system
  i686-linux</literal>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-attr"><term><option>--attr</option> / <option>-A</option>
<replaceable>attrPath</replaceable></term>

  <listitem><para>Select an attribute from the top-level Nix
  expression being evaluated.  (<command>nix-env</command>,
  <command>nix-instantiate</command>, <command>nix-build</command> and
  <command>nix-shell</command> only.)  The <emphasis>attribute
  path</emphasis> <replaceable>attrPath</replaceable> is a sequence of
  attribute names separated by dots.  For instance, given a top-level
  Nix expression <replaceable>e</replaceable>, the attribute path
  <literal>xorg.xorgserver</literal> would cause the expression
  <literal><replaceable>e</replaceable>.xorg.xorgserver</literal> to
  be used.  See <link linkend="refsec-nix-env-install-examples"><command>nix-env
  --install</command></link> for some concrete examples.</para>

  <para>In addition to attribute names, you can also specify array
  indices.  For instance, the attribute path
  <literal>foo.3.bar</literal> selects the <literal>bar</literal>
  attribute of the fourth element of the array in the
  <literal>foo</literal> attribute of the top-level
  expression.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--expr</option> / <option>-E</option></term>

  <listitem><para>Interpret the command line arguments as a list of
  Nix expressions to be parsed and evaluated, rather than as a list
  of file names of Nix expressions.
  (<command>nix-instantiate</command>, <command>nix-build</command>
  and <command>nix-shell</command> only.)</para>

  <para>For <command>nix-shell</command>, this option is commonly used
  to give you a shell in which you can build the packages returned
  by the expression. If you want to get a shell which contain the
  <emphasis>built</emphasis> packages ready for use, give your
  expression to the <command>nix-shell -p</command> convenience flag
  instead.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-I"><term><option>-I</option> <replaceable>path</replaceable></term>

  <listitem><para>Add a path to the Nix expression search path.  This
  option may be given multiple times.  See the <envar linkend="env-NIX_PATH">NIX_PATH</envar> environment variable for
  information on the semantics of the Nix search path.  Paths added
  through <option>-I</option> take precedence over
  <envar>NIX_PATH</envar>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--option</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>Set the Nix configuration option
  <replaceable>name</replaceable> to <replaceable>value</replaceable>.
  This overrides settings in the Nix configuration file (see
  <citerefentry><refentrytitle>nix.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--repair</option></term>

  <listitem><para>Fix corrupted or missing store paths by
  redownloading or rebuilding them.  Note that this is slow because it
  requires computing a cryptographic hash of the contents of every
  path in the closure of the build.  Also note the warning under
  <command>nix-store --repair-path</command>.</para></listitem>

</varlistentry>
</variablelist>

</refsection>


<refsection><title>Environment variables</title>

<variablelist>

  <varlistentry><term><envar>NIX_BUILD_SHELL</envar></term>

    <listitem><para>Shell used to start the interactive environment.
    Defaults to the <command>bash</command> found in <envar>PATH</envar>.</para></listitem>

  </varlistentry>

</variablelist>

</refsection>


<refsection><title>Examples</title>

<para>To build the dependencies of the package Pan, and start an
interactive shell in which to build it:

<screen>
$ nix-shell '&lt;nixpkgs&gt;' -A pan
[nix-shell]$ unpackPhase
[nix-shell]$ cd pan-*
[nix-shell]$ configurePhase
[nix-shell]$ buildPhase
[nix-shell]$ ./pan/gui/pan
</screen>

To clear the environment first, and do some additional automatic
initialisation of the interactive shell:

<screen>
$ nix-shell '&lt;nixpkgs&gt;' -A pan --pure \
    --command 'export NIX_DEBUG=1; export NIX_CORES=8; return'
</screen>

Nix expressions can also be given on the command line using the
<command>-E</command> and <command>-p</command> flags.
For instance, the following starts a shell containing the packages
<literal>sqlite</literal> and <literal>libX11</literal>:

<screen>
$ nix-shell -E 'with import &lt;nixpkgs&gt; { }; runCommand "dummy" { buildInputs = [ sqlite xorg.libX11 ]; } ""'
</screen>

A shorter way to do the same is:

<screen>
$ nix-shell -p sqlite xorg.libX11
[nix-shell]$ echo $NIX_LDFLAGS
&#x2026; -L/nix/store/j1zg5v&#x2026;-sqlite-3.8.0.2/lib -L/nix/store/0gmcz9&#x2026;-libX11-1.6.1/lib &#x2026;
</screen>

Note that <command>-p</command> accepts multiple full nix expressions that
are valid in the <literal>buildInputs = [ ... ]</literal> shown above,
not only package names. So the following is also legal:

<screen>
$ nix-shell -p sqlite 'git.override { withManual = false; }'
</screen>

The <command>-p</command> flag looks up Nixpkgs in the Nix search
path. You can override it by passing <option>-I</option> or setting
<envar>NIX_PATH</envar>. For example, the following gives you a shell
containing the Pan package from a specific revision of Nixpkgs:

<screen>
$ nix-shell -p pan -I nixpkgs=https://github.com/NixOS/nixpkgs/archive/8a3eea054838b55aca962c3fbde9c83c102b8bf2.tar.gz

[nix-shell:~]$ pan --version
Pan 0.139
</screen>

</para>

</refsection>


<refsection xml:id="ssec-nix-shell-shebang"><title>Use as a <literal>#!</literal>-interpreter</title>

<para>You can use <command>nix-shell</command> as a script interpreter
to allow scripts written in arbitrary languages to obtain their own
dependencies via Nix. This is done by starting the script with the
following lines:

<programlisting>
#! /usr/bin/env nix-shell
#! nix-shell -i <replaceable>real-interpreter</replaceable> -p <replaceable>packages</replaceable>
</programlisting>

where <replaceable>real-interpreter</replaceable> is the &#x201C;real&#x201D; script
interpreter that will be invoked by <command>nix-shell</command> after
it has obtained the dependencies and initialised the environment, and
<replaceable>packages</replaceable> are the attribute names of the
dependencies in Nixpkgs.</para>

<para>The lines starting with <literal>#! nix-shell</literal> specify
<command>nix-shell</command> options (see above). Note that you cannot
write <literal>#! /usr/bin/env nix-shell -i ...</literal> because
many operating systems only allow one argument in
<literal>#!</literal> lines.</para>

<para>For example, here is a Python script that depends on Python and
the <literal>prettytable</literal> package:

<programlisting>
#! /usr/bin/env nix-shell
#! nix-shell -i python -p python pythonPackages.prettytable

import prettytable

# Print a simple table.
t = prettytable.PrettyTable(["N", "N^2"])
for n in range(1, 10): t.add_row([n, n * n])
print t
</programlisting>

</para>

<para>Similarly, the following is a Perl script that specifies that it
requires Perl and the <literal>HTML::TokeParser::Simple</literal> and
<literal>LWP</literal> packages:

<programlisting>
#! /usr/bin/env nix-shell
#! nix-shell -i perl -p perl perlPackages.HTMLTokeParserSimple perlPackages.LWP

use HTML::TokeParser::Simple;

# Fetch nixos.org and print all hrefs.
my $p = HTML::TokeParser::Simple-&gt;new(url =&gt; 'http://nixos.org/');

while (my $token = $p-&gt;get_tag("a")) {
    my $href = $token-&gt;get_attr("href");
    print "$href\n" if $href;
}
</programlisting>

</para>

<para>Sometimes you need to pass a simple Nix expression to customize
a package like Terraform:

<programlisting><![CDATA[
#! /usr/bin/env nix-shell
#! nix-shell -i bash -p "terraform.withPlugins (plugins: [ plugins.openstack ])"

terraform apply
]]></programlisting>

<note><para>You must use double quotes (<literal>"</literal>) when
passing a simple Nix expression in a nix-shell shebang.</para></note>
</para>

<para>Finally, using the merging of multiple nix-shell shebangs the
following Haskell script uses a specific branch of Nixpkgs/NixOS (the
18.03 stable branch):

<programlisting><![CDATA[
#! /usr/bin/env nix-shell
#! nix-shell -i runghc -p "haskellPackages.ghcWithPackages (ps: [ps.HTTP ps.tagsoup])"
#! nix-shell -I nixpkgs=https://github.com/NixOS/nixpkgs/archive/nixos-18.03.tar.gz

import Network.HTTP
import Text.HTML.TagSoup

-- Fetch nixos.org and print all hrefs.
main = do
  resp <- Network.HTTP.simpleHTTP (getRequest "http://nixos.org/")
  body <- getResponseBody resp
  let tags = filter (isTagOpenName "a") $ parseTags body
  let tags' = map (fromAttrib "href") tags
  mapM_ putStrLn $ filter (/= "") tags'
]]></programlisting>

If you want to be even more precise, you can specify a specific
revision of Nixpkgs:

<programlisting>
#! nix-shell -I nixpkgs=https://github.com/NixOS/nixpkgs/archive/0672315759b3e15e2121365f067c1c8c56bb4722.tar.gz
</programlisting>

</para>

<para>The examples above all used <option>-p</option> to get
dependencies from Nixpkgs. You can also use a Nix expression to build
your own dependencies. For example, the Python example could have been
written as:

<programlisting>
#! /usr/bin/env nix-shell
#! nix-shell deps.nix -i python
</programlisting>

where the file <filename>deps.nix</filename> in the same directory
as the <literal>#!</literal>-script contains:

<programlisting>
with import &lt;nixpkgs&gt; {};

runCommand "dummy" { buildInputs = [ python pythonPackages.prettytable ]; } ""
</programlisting>

</para>

</refsection>


<refsection condition="manpage"><title>Environment variables</title>

<variablelist>
  <varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>IN_NIX_SHELL</envar></term>

  <listitem><para>Indicator that tells if the current environment was set up by
  <command>nix-shell</command>.  Since Nix 2.0 the values are
  <literal>"pure"</literal> and <literal>"impure"</literal></para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="env-NIX_PATH"><term><envar>NIX_PATH</envar></term>

  <listitem>

    <para>A colon-separated list of directories used to look up Nix
    expressions enclosed in angle brackets (i.e.,
    <literal>&lt;<replaceable>path</replaceable>&gt;</literal>).  For
    instance, the value

    <screen>
/home/eelco/Dev:/etc/nixos</screen>

    will cause Nix to look for paths relative to
    <filename>/home/eelco/Dev</filename> and
    <filename>/etc/nixos</filename>, in this order.  It is also
    possible to match paths against a prefix.  For example, the value

    <screen>
nixpkgs=/home/eelco/Dev/nixpkgs-branch:/etc/nixos</screen>

    will cause Nix to search for
    <literal>&lt;nixpkgs/<replaceable>path</replaceable>&gt;</literal> in
    <filename>/home/eelco/Dev/nixpkgs-branch/<replaceable>path</replaceable></filename>
    and
    <filename>/etc/nixos/nixpkgs/<replaceable>path</replaceable></filename>.</para>

    <para>If a path in the Nix search path starts with
    <literal>http://</literal> or <literal>https://</literal>, it is
    interpreted as the URL of a tarball that will be downloaded and
    unpacked to a temporary location. The tarball must consist of a
    single top-level directory. For example, setting
    <envar>NIX_PATH</envar> to

    <screen>
nixpkgs=https://github.com/NixOS/nixpkgs/archive/nixos-15.09.tar.gz</screen>

    tells Nix to download the latest revision in the Nixpkgs/NixOS
    15.09 channel.</para>

    <para>A following shorthand can be used to refer to the official channels:

    <screen>nixpkgs=channel:nixos-15.09</screen>
    </para>

    <para>The search path can be extended using the <option linkend="opt-I">-I</option> option, which takes precedence over
    <envar>NIX_PATH</envar>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_IGNORE_SYMLINK_STORE</envar></term>

  <listitem>

  <para>Normally, the Nix store directory (typically
  <filename>/nix/store</filename>) is not allowed to contain any
  symlink components.  This is to prevent &#x201C;impure&#x201D; builds.  Builders
  sometimes &#x201C;canonicalise&#x201D; paths by resolving all symlink components.
  Thus, builds on different machines (with
  <filename>/nix/store</filename> resolving to different locations)
  could yield different results.  This is generally not a problem,
  except when builds are deployed to machines where
  <filename>/nix/store</filename> resolves differently.  If you are
  sure that you&#x2019;re not going to do that, you can set
  <envar>NIX_IGNORE_SYMLINK_STORE</envar> to <envar>1</envar>.</para>

  <para>Note that if you&#x2019;re symlinking the Nix store so that you can
  put it on another file system than the root file system, on Linux
  you&#x2019;re better off using <literal>bind</literal> mount points, e.g.,

  <screen>
$ mkdir /nix
$ mount -o bind /mnt/otherdisk/nix /nix</screen>

  Consult the <citerefentry><refentrytitle>mount</refentrytitle>
  <manvolnum>8</manvolnum></citerefentry> manual page for details.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_STORE_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix store (default
  <filename><replaceable>prefix</replaceable>/store</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_DATA_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix static data
  directory (default
  <filename><replaceable>prefix</replaceable>/share</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_LOG_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix log directory
  (default <filename><replaceable>prefix</replaceable>/var/log/nix</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_STATE_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix state directory
  (default <filename><replaceable>prefix</replaceable>/var/nix</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_CONF_DIR</envar></term>

  <listitem><para>Overrides the location of the system Nix configuration
  directory (default
  <filename><replaceable>prefix</replaceable>/etc/nix</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_USER_CONF_FILES</envar></term>

  <listitem><para>Overrides the location of the user Nix configuration files
  to load from (defaults to the XDG spec locations). The variable is treated
  as a list separated by the <literal>:</literal> token.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>TMPDIR</envar></term>

  <listitem><para>Use the specified directory to store temporary
  files.  In particular, this includes temporary build directories;
  these can take up substantial amounts of disk space.  The default is
  <filename>/tmp</filename>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="envar-remote"><term><envar>NIX_REMOTE</envar></term>

  <listitem><para>This variable should be set to
  <literal>daemon</literal> if you want to use the Nix daemon to
  execute Nix operations. This is necessary in <link linkend="ssec-multi-user">multi-user Nix installations</link>.
  If the Nix daemon's Unix socket is at some non-standard path,
  this variable should be set to <literal>unix://path/to/socket</literal>.
  Otherwise, it should be left unset.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_SHOW_STATS</envar></term>

  <listitem><para>If set to <literal>1</literal>, Nix will print some
  evaluation statistics, such as the number of values
  allocated.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_COUNT_CALLS</envar></term>

  <listitem><para>If set to <literal>1</literal>, Nix will print how
  often functions were called during Nix expression evaluation.  This
  is useful for profiling your Nix expressions.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>GC_INITIAL_HEAP_SIZE</envar></term>

  <listitem><para>If Nix has been configured to use the Boehm garbage
  collector, this variable sets the initial size of the heap in bytes.
  It defaults to 384 MiB.  Setting it to a low value reduces memory
  consumption, but will increase runtime due to the overhead of
  garbage collection.</para></listitem>

</varlistentry>
</variablelist>

</refsection>


</refentry>
<refentry xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-nix-store">

<refmeta>
  <refentrytitle>nix-store</refentrytitle>
  <manvolnum>1</manvolnum>
  <refmiscinfo class="source">Nix</refmiscinfo>
  <refmiscinfo class="version">3.0</refmiscinfo>
</refmeta>

<refnamediv>
  <refname>nix-store</refname>
  <refpurpose>manipulate or query the Nix store</refpurpose>
</refnamediv>

<refsynopsisdiv>
  <cmdsynopsis>
    <command>nix-store</command>
    <arg xmlns="http://docbook.org/ns/docbook"><option>--help</option></arg><arg xmlns="http://docbook.org/ns/docbook"><option>--version</option></arg><arg xmlns="http://docbook.org/ns/docbook" rep="repeat">
  <group choice="req">
    <arg choice="plain"><option>--verbose</option></arg>
    <arg choice="plain"><option>-v</option></arg>
  </group>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <arg choice="plain"><option>--quiet</option></arg>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>--log-format</option>
  <replaceable>format</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <group choice="plain">
    <arg choice="plain"><option>--no-build-output</option></arg>
    <arg choice="plain"><option>-Q</option></arg>
  </group>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <group choice="req">
    <arg choice="plain"><option>--max-jobs</option></arg>
    <arg choice="plain"><option>-j</option></arg>
  </group>
  <replaceable>number</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>--cores</option>
  <replaceable>number</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>--max-silent-time</option>
  <replaceable>number</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>--timeout</option>
  <replaceable>number</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <group choice="plain">
    <arg choice="plain"><option>--keep-going</option></arg>
    <arg choice="plain"><option>-k</option></arg>
  </group>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <group choice="plain">
    <arg choice="plain"><option>--keep-failed</option></arg>
    <arg choice="plain"><option>-K</option></arg>
  </group>
</arg><arg xmlns="http://docbook.org/ns/docbook"><option>--fallback</option></arg><arg xmlns="http://docbook.org/ns/docbook"><option>--readonly-mode</option></arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>-I</option>
  <replaceable>path</replaceable>
</arg><arg xmlns="http://docbook.org/ns/docbook">
  <option>--option</option>
  <replaceable>name</replaceable>
  <replaceable>value</replaceable>
</arg><sbr xmlns="http://docbook.org/ns/docbook"/>
    <arg><option>--add-root</option> <replaceable>path</replaceable></arg>
    <arg><option>--indirect</option></arg>
    <arg choice="plain"><replaceable>operation</replaceable></arg>
    <arg rep="repeat"><replaceable>options</replaceable></arg>
    <arg rep="repeat"><replaceable>arguments</replaceable></arg>
  </cmdsynopsis>
</refsynopsisdiv>


<refsection><title>Description</title>

<para>The command <command>nix-store</command> performs primitive
operations on the Nix store.  You generally do not need to run this
command manually.</para>

<para><command>nix-store</command> takes exactly one
<emphasis>operation</emphasis> flag which indicates the subcommand to
be performed.  These are documented below.</para>

</refsection>



<!--######################################################################-->

<refsection><title>Common options</title>

<para>This section lists the options that are common to all
operations.  These options are allowed for every subcommand, though
they may not always have an effect.  <phrase condition="manual">See
also <xref linkend="sec-common-options"/> for a list of common
options.</phrase></para>

<variablelist>

  <varlistentry xml:id="opt-add-root"><term><option>--add-root</option> <replaceable>path</replaceable></term>

    <listitem><para>Causes the result of a realisation
    (<option>--realise</option> and <option>--force-realise</option>)
    to be registered as a root of the garbage collector<phrase condition="manual"> (see <xref linkend="ssec-gc-roots"/>)</phrase>.  The root is stored in
    <replaceable>path</replaceable>, which must be inside a directory
    that is scanned for roots by the garbage collector (i.e.,
    typically in a subdirectory of
    <filename>/nix/var/nix/gcroots/</filename>)
    <emphasis>unless</emphasis> the <option>--indirect</option> flag
    is used.</para>

    <para>If there are multiple results, then multiple symlinks will
    be created by sequentially numbering symlinks beyond the first one
    (e.g., <filename>foo</filename>, <filename>foo-2</filename>,
    <filename>foo-3</filename>, and so on).</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--indirect</option></term>

    <listitem>

    <para>In conjunction with <option>--add-root</option>, this option
    allows roots to be stored <emphasis>outside</emphasis> of the GC
    roots directory.  This is useful for commands such as
    <command>nix-build</command> that place a symlink to the build
    result in the current directory; such a build result should not be
    garbage-collected unless the symlink is removed.</para>

    <para>The <option>--indirect</option> flag causes a uniquely named
    symlink to <replaceable>path</replaceable> to be stored in
    <filename>/nix/var/nix/gcroots/auto/</filename>.  For instance,

    <screen>
$ nix-store --add-root /home/eelco/bla/result --indirect -r <replaceable>...</replaceable>

$ ls -l /nix/var/nix/gcroots/auto
lrwxrwxrwx    1 ... 2005-03-13 21:10 dn54lcypm8f8... -&gt; /home/eelco/bla/result

$ ls -l /home/eelco/bla/result
lrwxrwxrwx    1 ... 2005-03-13 21:10 /home/eelco/bla/result -&gt; /nix/store/1r11343n6qd4...-f-spot-0.0.10</screen>

    Thus, when <filename>/home/eelco/bla/result</filename> is removed,
    the GC root in the <filename>auto</filename> directory becomes a
    dangling symlink and will be ignored by the collector.</para>

    <warning><para>Note that it is not possible to move or rename
    indirect GC roots, since the symlink in the
    <filename>auto</filename> directory will still point to the old
    location.</para></warning>

    </listitem>

  </varlistentry>

</variablelist>

<variablelist condition="manpage">
  <varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--help</option></term>

  <listitem><para>Prints out a summary of the command syntax and
  exits.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--version</option></term>

  <listitem><para>Prints out the Nix version number on standard output
  and exits.</para></listitem>
</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--verbose</option> / <option>-v</option></term>

  <listitem>

  <para>Increases the level of verbosity of diagnostic messages
  printed on standard error.  For each Nix operation, the information
  printed on standard output is well-defined; any diagnostic
  information is printed on standard error, never on standard
  output.</para>

  <para>This option may be specified repeatedly.  Currently, the
  following verbosity levels exist:</para>

  <variablelist>

    <varlistentry><term>0</term>
    <listitem><para>&#x201C;Errors only&#x201D;: only print messages
    explaining why the Nix invocation failed.</para></listitem>
    </varlistentry>

    <varlistentry><term>1</term>
    <listitem><para>&#x201C;Informational&#x201D;: print
    <emphasis>useful</emphasis> messages about what Nix is doing.
    This is the default.</para></listitem>
    </varlistentry>

    <varlistentry><term>2</term>
    <listitem><para>&#x201C;Talkative&#x201D;: print more informational
    messages.</para></listitem>
    </varlistentry>

    <varlistentry><term>3</term>
    <listitem><para>&#x201C;Chatty&#x201D;: print even more
    informational messages.</para></listitem>
    </varlistentry>

    <varlistentry><term>4</term>
    <listitem><para>&#x201C;Debug&#x201D;: print debug
    information.</para></listitem>
    </varlistentry>

    <varlistentry><term>5</term>
    <listitem><para>&#x201C;Vomit&#x201D;: print vast amounts of debug
    information.</para></listitem>
    </varlistentry>

  </variablelist>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--quiet</option></term>

  <listitem>

  <para>Decreases the level of verbosity of diagnostic messages
  printed on standard error.  This is the inverse option to
  <option>-v</option> / <option>--verbose</option>.
  </para>

  <para>This option may be specified repeatedly.  See the previous
  verbosity levels list.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-log-format"><term><option>--log-format</option> <replaceable>format</replaceable></term>

  <listitem>

  <para>This option can be used to change the output of the log format, with
  <replaceable>format</replaceable> being one of:</para>

  <variablelist>

    <varlistentry><term>raw</term>
    <listitem><para>This is the raw format, as outputted by nix-build.</para></listitem>
    </varlistentry>

    <varlistentry><term>internal-json</term>
    <listitem><para>Outputs the logs in a structured manner. NOTE: the json schema is not guarantees to be stable between releases.</para></listitem>
    </varlistentry>

    <varlistentry><term>bar</term>
    <listitem><para>Only display a progress bar during the builds.</para></listitem>
    </varlistentry>

    <varlistentry><term>bar-with-logs</term>
    <listitem><para>Display the raw logs, with the progress bar at the bottom.</para></listitem>
    </varlistentry>

  </variablelist>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--no-build-output</option> / <option>-Q</option></term>

  <listitem><para>By default, output written by builders to standard
  output and standard error is echoed to the Nix command's standard
  error.  This option suppresses this behaviour.  Note that the
  builder's standard output and error are always written to a log file
  in
  <filename><replaceable>prefix</replaceable>/nix/var/log/nix</filename>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-max-jobs"><term><option>--max-jobs</option> / <option>-j</option>
<replaceable>number</replaceable></term>

  <listitem>

  <para>Sets the maximum number of build jobs that Nix will
  perform in parallel to the specified number.  Specify
  <literal>auto</literal> to use the number of CPUs in the system.
  The default is specified by the <link linkend="conf-max-jobs"><literal>max-jobs</literal></link>
  configuration setting, which itself defaults to
  <literal>1</literal>.  A higher value is useful on SMP systems or to
  exploit I/O latency.</para>

  <para> Setting it to <literal>0</literal> disallows building on the local
  machine, which is useful when you want builds to happen only on remote
  builders.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-cores"><term><option>--cores</option></term>

  <listitem><para>Sets the value of the <envar>NIX_BUILD_CORES</envar>
  environment variable in the invocation of builders.  Builders can
  use this variable at their discretion to control the maximum amount
  of parallelism.  For instance, in Nixpkgs, if the derivation
  attribute <varname>enableParallelBuilding</varname> is set to
  <literal>true</literal>, the builder passes the
  <option>-j<replaceable>N</replaceable></option> flag to GNU Make.
  It defaults to the value of the <link linkend="conf-cores"><literal>cores</literal></link>
  configuration setting, if set, or <literal>1</literal> otherwise.
  The value <literal>0</literal> means that the builder should use all
  available CPU cores in the system.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-max-silent-time"><term><option>--max-silent-time</option></term>

  <listitem><para>Sets the maximum number of seconds that a builder
  can go without producing any data on standard output or standard
  error.  The default is specified by the <link linkend="conf-max-silent-time"><literal>max-silent-time</literal></link>
  configuration setting.  <literal>0</literal> means no
  time-out.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-timeout"><term><option>--timeout</option></term>

  <listitem><para>Sets the maximum number of seconds that a builder
  can run.  The default is specified by the <link linkend="conf-timeout"><literal>timeout</literal></link>
  configuration setting.  <literal>0</literal> means no
  timeout.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--keep-going</option> / <option>-k</option></term>

  <listitem><para>Keep going in case of failed builds, to the
  greatest extent possible.  That is, if building an input of some
  derivation fails, Nix will still build the other inputs, but not the
  derivation itself.  Without this option, Nix stops if any build
  fails (except for builds of substitutes), possibly killing builds in
  progress (in case of parallel or distributed builds).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--keep-failed</option> / <option>-K</option></term>

  <listitem><para>Specifies that in case of a build failure, the
  temporary directory (usually in <filename>/tmp</filename>) in which
  the build takes place should not be deleted.  The path of the build
  directory is printed as an informational message.
    </para>
  </listitem>
</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--fallback</option></term>

  <listitem>

  <para>Whenever Nix attempts to build a derivation for which
  substitutes are known for each output path, but realising the output
  paths through the substitutes fails, fall back on building the
  derivation.</para>

  <para>The most common scenario in which this is useful is when we
  have registered substitutes in order to perform binary distribution
  from, say, a network repository.  If the repository is down, the
  realisation of the derivation will fail.  When this option is
  specified, Nix will build the derivation instead.  Thus,
  installation from binaries falls back on installation from source.
  This option is not the default since it is generally not desirable
  for a transient failure in obtaining the substitutes to lead to a
  full build from source (with the related consumption of
  resources).</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--no-build-hook</option></term>

  <listitem>

  <para>Disables the build hook mechanism.  This allows to ignore remote
  builders if they are setup on the machine.</para>

  <para>It's useful in cases where the bandwidth between the client and the
  remote builder is too low.  In that case it can take more time to upload the
  sources to the remote builder and fetch back the result than to do the
  computation locally.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--readonly-mode</option></term>

  <listitem><para>When this option is used, no attempt is made to open
  the Nix database.  Most Nix operations do need database access, so
  those operations will fail.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--arg</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>This option is accepted by
  <command>nix-env</command>, <command>nix-instantiate</command>,
  <command>nix-shell</command> and <command>nix-build</command>.
  When evaluating Nix expressions, the expression evaluator will
  automatically try to call functions that
  it encounters.  It can automatically call functions for which every
  argument has a <link linkend="ss-functions">default value</link>
  (e.g., <literal>{ <replaceable>argName</replaceable> ?
  <replaceable>defaultValue</replaceable> }:
  <replaceable>...</replaceable></literal>).  With
  <option>--arg</option>, you can also call functions that have
  arguments without a default value (or override a default value).
  That is, if the evaluator encounters a function with an argument
  named <replaceable>name</replaceable>, it will call it with value
  <replaceable>value</replaceable>.</para>

  <para>For instance, the top-level <literal>default.nix</literal> in
  Nixpkgs is actually a function:

<programlisting>
{ # The system (e.g., `i686-linux') for which to build the packages.
  system ? builtins.currentSystem
  <replaceable>...</replaceable>
}: <replaceable>...</replaceable></programlisting>

  So if you call this Nix expression (e.g., when you do
  <literal>nix-env -i <replaceable>pkgname</replaceable></literal>),
  the function will be called automatically using the value <link linkend="builtin-currentSystem"><literal>builtins.currentSystem</literal></link>
  for the <literal>system</literal> argument.  You can override this
  using <option>--arg</option>, e.g., <literal>nix-env -i
  <replaceable>pkgname</replaceable> --arg system
  \"i686-freebsd\"</literal>.  (Note that since the argument is a Nix
  string literal, you have to escape the quotes.)</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--argstr</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>This option is like <option>--arg</option>, only the
  value is not a Nix expression but a string.  So instead of
  <literal>--arg system \"i686-linux\"</literal> (the outer quotes are
  to keep the shell happy) you can say <literal>--argstr system
  i686-linux</literal>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-attr"><term><option>--attr</option> / <option>-A</option>
<replaceable>attrPath</replaceable></term>

  <listitem><para>Select an attribute from the top-level Nix
  expression being evaluated.  (<command>nix-env</command>,
  <command>nix-instantiate</command>, <command>nix-build</command> and
  <command>nix-shell</command> only.)  The <emphasis>attribute
  path</emphasis> <replaceable>attrPath</replaceable> is a sequence of
  attribute names separated by dots.  For instance, given a top-level
  Nix expression <replaceable>e</replaceable>, the attribute path
  <literal>xorg.xorgserver</literal> would cause the expression
  <literal><replaceable>e</replaceable>.xorg.xorgserver</literal> to
  be used.  See <link linkend="refsec-nix-env-install-examples"><command>nix-env
  --install</command></link> for some concrete examples.</para>

  <para>In addition to attribute names, you can also specify array
  indices.  For instance, the attribute path
  <literal>foo.3.bar</literal> selects the <literal>bar</literal>
  attribute of the fourth element of the array in the
  <literal>foo</literal> attribute of the top-level
  expression.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--expr</option> / <option>-E</option></term>

  <listitem><para>Interpret the command line arguments as a list of
  Nix expressions to be parsed and evaluated, rather than as a list
  of file names of Nix expressions.
  (<command>nix-instantiate</command>, <command>nix-build</command>
  and <command>nix-shell</command> only.)</para>

  <para>For <command>nix-shell</command>, this option is commonly used
  to give you a shell in which you can build the packages returned
  by the expression. If you want to get a shell which contain the
  <emphasis>built</emphasis> packages ready for use, give your
  expression to the <command>nix-shell -p</command> convenience flag
  instead.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-I"><term><option>-I</option> <replaceable>path</replaceable></term>

  <listitem><para>Add a path to the Nix expression search path.  This
  option may be given multiple times.  See the <envar linkend="env-NIX_PATH">NIX_PATH</envar> environment variable for
  information on the semantics of the Nix search path.  Paths added
  through <option>-I</option> take precedence over
  <envar>NIX_PATH</envar>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--option</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>Set the Nix configuration option
  <replaceable>name</replaceable> to <replaceable>value</replaceable>.
  This overrides settings in the Nix configuration file (see
  <citerefentry><refentrytitle>nix.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--repair</option></term>

  <listitem><para>Fix corrupted or missing store paths by
  redownloading or rebuilding them.  Note that this is slow because it
  requires computing a cryptographic hash of the contents of every
  path in the closure of the build.  Also note the warning under
  <command>nix-store --repair-path</command>.</para></listitem>

</varlistentry>
</variablelist>

</refsection>



<!--######################################################################-->

<refsection xml:id="rsec-nix-store-realise"><title>Operation <option>--realise</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-store</command>
  <group choice="req">
    <arg choice="plain"><option>--realise</option></arg>
    <arg choice="plain"><option>-r</option></arg>
  </group>
  <arg choice="plain" rep="repeat"><replaceable>paths</replaceable></arg>
  <arg><option>--dry-run</option></arg>
</cmdsynopsis>

</refsection>

<refsection><title>Description</title>

<para>The operation <option>--realise</option> essentially &#x201C;builds&#x201D;
the specified store paths.  Realisation is a somewhat overloaded term:

<itemizedlist>

  <listitem><para>If the store path is a
  <emphasis>derivation</emphasis>, realisation ensures that the output
  paths of the derivation are <link linkend="gloss-validity">valid</link> (i.e., the output path and its
  closure exist in the file system).  This can be done in several
  ways.  First, it is possible that the outputs are already valid, in
  which case we are done immediately.  Otherwise, there may be <link linkend="gloss-substitute">substitutes</link> that produce the
  outputs (e.g., by downloading them).  Finally, the outputs can be
  produced by performing the build action described by the
  derivation.</para></listitem>

  <listitem><para>If the store path is not a derivation, realisation
  ensures that the specified path is valid (i.e., it and its closure
  exist in the file system).  If the path is already valid, we are
  done immediately.  Otherwise, the path and any missing paths in its
  closure may be produced through substitutes.  If there are no
  (successful) subsitutes, realisation fails.</para></listitem>

</itemizedlist>

</para>

<para>The output path of each derivation is printed on standard
output.  (For non-derivations argument, the argument itself is
printed.)</para>

<para>The following flags are available:</para>

<variablelist>

  <varlistentry><term><option>--dry-run</option></term>

    <listitem><para>Print on standard error a description of what
    packages would be built or downloaded, without actually performing
    the operation.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--ignore-unknown</option></term>

    <listitem><para>If a non-derivation path does not have a
    substitute, then silently ignore it.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--check</option></term>

    <listitem><para>This option allows you to check whether a
    derivation is deterministic. It rebuilds the specified derivation
    and checks whether the result is bitwise-identical with the
    existing outputs, printing an error if that&#x2019;s not the case. The
    outputs of the specified derivation must already exist. When used
    with <option>-K</option>, if an output path is not identical to
    the corresponding output from the previous build, the new output
    path is left in
    <filename>/nix/store/<replaceable>name</replaceable>.check.</filename></para>

    <para>See also the <option>build-repeat</option> configuration
    option, which repeats a derivation a number of times and prevents
    its outputs from being registered as &#x201C;valid&#x201D; in the Nix store
    unless they are identical.</para></listitem>

  </varlistentry>

</variablelist>

<para>Special exit codes:</para>

<variablelist>

  <varlistentry><term><literal>100</literal></term>
    <listitem><para>Generic build failure, the builder process
    returned with a non-zero exit code.</para></listitem>
  </varlistentry>

  <varlistentry><term><literal>101</literal></term>
    <listitem><para>Build timeout, the build was aborted because it
    did not complete within the specified <link linkend="conf-timeout"><literal>timeout</literal></link>.
    </para></listitem>
  </varlistentry>

  <varlistentry><term><literal>102</literal></term>
    <listitem><para>Hash mismatch, the build output was rejected
    because it does not match the specified <link linkend="fixed-output-drvs"><varname>outputHash</varname></link>.
    </para></listitem>
  </varlistentry>

  <varlistentry><term><literal>104</literal></term>
    <listitem><para>Not deterministic, the build succeeded in check
    mode but the resulting output is not binary reproducable.</para>
    </listitem>
  </varlistentry>

</variablelist>

<para>With the <option>--keep-going</option> flag it's possible for
multiple failures to occur, in this case the 1xx status codes are or combined
using binary or. <screen>
1100100
   ^^^^
   |||`- timeout
   ||`-- output hash mismatch
   |`--- build failure
   `---- not deterministic
</screen></para>

</refsection>


<refsection><title>Examples</title>

<para>This operation is typically used to build store derivations
produced by <link linkend="sec-nix-instantiate"><command>nix-instantiate</command></link>:

<screen>
$ nix-store -r $(nix-instantiate ./test.nix)
/nix/store/31axcgrlbfsxzmfff1gyj1bf62hvkby2-aterm-2.3.1</screen>

This is essentially what <link linkend="sec-nix-build"><command>nix-build</command></link> does.</para>

<para>To test whether a previously-built derivation is deterministic:

<screen>
$ nix-build '&lt;nixpkgs&gt;' -A hello --check -K
</screen>

</para>

</refsection>


</refsection>



<!--######################################################################-->

<refsection xml:id="rsec-nix-store-serve"><title>Operation <option>--serve</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-store</command>
  <arg choice="plain"><option>--serve</option></arg>
  <arg><option>--write</option></arg>
</cmdsynopsis>

</refsection>

<refsection><title>Description</title>

<para>The operation <option>--serve</option> provides access to
the Nix store over stdin and stdout, and is intended to be used
as a means of providing Nix store access to a restricted ssh user.
</para>

<para>The following flags are available:</para>

<variablelist>

  <varlistentry><term><option>--write</option></term>

    <listitem><para>Allow the connected client to request the realization
    of derivations. In effect, this can be used to make the host act
    as a remote builder.</para></listitem>

  </varlistentry>

</variablelist>

</refsection>


<refsection><title>Examples</title>

<para>To turn a host into a build server, the
<filename>authorized_keys</filename> file can be used to provide build
access to a given SSH public key:

<screen>
$ cat &lt;&lt;EOF &gt;&gt;/root/.ssh/authorized_keys
command="nice -n20 nix-store --serve --write" ssh-rsa AAAAB3NzaC1yc2EAAAA...
EOF
</screen>

</para>

</refsection>


</refsection>



<!--######################################################################-->

<refsection xml:id="rsec-nix-store-gc"><title>Operation <option>--gc</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-store</command>
  <arg choice="plain"><option>--gc</option></arg>
  <group>
    <arg choice="plain"><option>--print-roots</option></arg>
    <arg choice="plain"><option>--print-live</option></arg>
    <arg choice="plain"><option>--print-dead</option></arg>
  </group>
  <arg><option>--max-freed</option> <replaceable>bytes</replaceable></arg>
</cmdsynopsis>

</refsection>

<refsection><title>Description</title>

<para>Without additional flags, the operation <option>--gc</option>
performs a garbage collection on the Nix store.  That is, all paths in
the Nix store not reachable via file system references from a set of
&#x201C;roots&#x201D;, are deleted.</para>

<para>The following suboperations may be specified:</para>

<variablelist>

  <varlistentry><term><option>--print-roots</option></term>

    <listitem><para>This operation prints on standard output the set
    of roots used by the garbage collector.  What constitutes a root
    is described in <xref linkend="ssec-gc-roots"/>.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--print-live</option></term>

    <listitem><para>This operation prints on standard output the set
    of &#x201C;live&#x201D; store paths, which are all the store paths reachable
    from the roots.  Live paths should never be deleted, since that
    would break consistency &#x2014; it would become possible that
    applications are installed that reference things that are no
    longer present in the store.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--print-dead</option></term>

    <listitem><para>This operation prints out on standard output the
    set of &#x201C;dead&#x201D; store paths, which is just the opposite of the set
    of live paths: any path in the store that is not live (with
    respect to the roots) is dead.</para></listitem>

  </varlistentry>

</variablelist>

<para>By default, all unreachable paths are deleted.  The following
options control what gets deleted and in what order:

<variablelist>

  <varlistentry><term><option>--max-freed</option> <replaceable>bytes</replaceable></term>

    <listitem><para>Keep deleting paths until at least
    <replaceable>bytes</replaceable> bytes have been deleted, then
    stop.  The argument <replaceable>bytes</replaceable> can be
    followed by the multiplicative suffix <literal>K</literal>,
    <literal>M</literal>, <literal>G</literal> or
    <literal>T</literal>, denoting KiB, MiB, GiB or TiB
    units.</para></listitem>

  </varlistentry>

</variablelist>

</para>

<para>The behaviour of the collector is also influenced by the <link linkend="conf-keep-outputs"><literal>keep-outputs</literal></link>
and <link linkend="conf-keep-derivations"><literal>keep-derivations</literal></link>
variables in the Nix configuration file.</para>

<para>By default, the collector prints the total number of freed bytes
when it finishes (or when it is interrupted). With
<option>--print-dead</option>, it prints the number of bytes that would
be freed.</para>

</refsection>


<refsection><title>Examples</title>

<para>To delete all unreachable paths, just do:

<screen>
$ nix-store --gc
deleting `/nix/store/kq82idx6g0nyzsp2s14gfsc38npai7lf-cairo-1.0.4.tar.gz.drv'
<replaceable>...</replaceable>
8825586 bytes freed (8.42 MiB)</screen>

</para>

<para>To delete at least 100 MiBs of unreachable paths:

<screen>
$ nix-store --gc --max-freed $((100 * 1024 * 1024))</screen>

</para>

</refsection>


</refsection>



<!--######################################################################-->

<refsection><title>Operation <option>--delete</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-store</command>
  <arg choice="plain"><option>--delete</option></arg>
  <arg><option>--ignore-liveness</option></arg>
  <arg choice="plain" rep="repeat"><replaceable>paths</replaceable></arg>
</cmdsynopsis>

</refsection>

<refsection><title>Description</title>

<para>The operation <option>--delete</option> deletes the store paths
<replaceable>paths</replaceable> from the Nix store, but only if it is
safe to do so; that is, when the path is not reachable from a root of
the garbage collector.  This means that you can only delete paths that
would also be deleted by <literal>nix-store --gc</literal>.  Thus,
<literal>--delete</literal> is a more targeted version of
<literal>--gc</literal>.</para>

<para>With the option <option>--ignore-liveness</option>, reachability
from the roots is ignored.  However, the path still won&#x2019;t be deleted
if there are other paths in the store that refer to it (i.e., depend
on it).</para>

</refsection>

<refsection><title>Example</title>

<screen>
$ nix-store --delete /nix/store/zq0h41l75vlb4z45kzgjjmsjxvcv1qk7-mesa-6.4
0 bytes freed (0.00 MiB)
error: cannot delete path `/nix/store/zq0h41l75vlb4z45kzgjjmsjxvcv1qk7-mesa-6.4' since it is still alive</screen>

</refsection>

</refsection>



<!--######################################################################-->

<refsection xml:id="refsec-nix-store-query"><title>Operation <option>--query</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-store</command>
  <group choice="req">
    <arg choice="plain"><option>--query</option></arg>
    <arg choice="plain"><option>-q</option></arg>
  </group>
  <group choice="req">
    <arg choice="plain"><option>--outputs</option></arg>
    <arg choice="plain"><option>--requisites</option></arg>
    <arg choice="plain"><option>-R</option></arg>
    <arg choice="plain"><option>--references</option></arg>
    <arg choice="plain"><option>--referrers</option></arg>
    <arg choice="plain"><option>--referrers-closure</option></arg>
    <arg choice="plain"><option>--deriver</option></arg>
    <arg choice="plain"><option>-d</option></arg>
    <arg choice="plain"><option>--graph</option></arg>
    <arg choice="plain"><option>--tree</option></arg>
    <arg choice="plain"><option>--binding</option> <replaceable>name</replaceable></arg>
    <arg choice="plain"><option>-b</option> <replaceable>name</replaceable></arg>
    <arg choice="plain"><option>--hash</option></arg>
    <arg choice="plain"><option>--size</option></arg>
    <arg choice="plain"><option>--roots</option></arg>
  </group>
  <arg><option>--use-output</option></arg>
  <arg><option>-u</option></arg>
  <arg><option>--force-realise</option></arg>
  <arg><option>-f</option></arg>
  <arg choice="plain" rep="repeat"><replaceable>paths</replaceable></arg>
</cmdsynopsis>

</refsection>


<refsection><title>Description</title>

<para>The operation <option>--query</option> displays various bits of
information about the store paths .  The queries are described below.  At
most one query can be specified.  The default query is
<option>--outputs</option>.</para>

<para>The paths <replaceable>paths</replaceable> may also be symlinks
from outside of the Nix store, to the Nix store.  In that case, the
query is applied to the target of the symlink.</para>


</refsection>


<refsection><title>Common query options</title>

<variablelist>

  <varlistentry><term><option>--use-output</option></term>
    <term><option>-u</option></term>

    <listitem><para>For each argument to the query that is a store
    derivation, apply the query to the output path of the derivation
    instead.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--force-realise</option></term>
    <term><option>-f</option></term>

    <listitem><para>Realise each argument to the query first (see
    <link linkend="rsec-nix-store-realise"><command>nix-store
    --realise</command></link>).</para></listitem>

  </varlistentry>

</variablelist>

</refsection>


<refsection xml:id="nixref-queries"><title>Queries</title>

<variablelist>

  <varlistentry><term><option>--outputs</option></term>

    <listitem><para>Prints out the <link linkend="gloss-output-path">output paths</link> of the store
    derivations <replaceable>paths</replaceable>.  These are the paths
    that will be produced when the derivation is
    built.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--requisites</option></term>
    <term><option>-R</option></term>

    <listitem><para>Prints out the <link linkend="gloss-closure">closure</link> of the store path
    <replaceable>paths</replaceable>.</para>

    <para>This query has one option:</para>

    <variablelist>

      <varlistentry><term><option>--include-outputs</option></term>

        <listitem><para>Also include the output path of store
        derivations, and their closures.</para></listitem>

      </varlistentry>

    </variablelist>

    <para>This query can be used to implement various kinds of
    deployment.  A <emphasis>source deployment</emphasis> is obtained
    by distributing the closure of a store derivation.  A
    <emphasis>binary deployment</emphasis> is obtained by distributing
    the closure of an output path.  A <emphasis>cache
    deployment</emphasis> (combined source/binary deployment,
    including binaries of build-time-only dependencies) is obtained by
    distributing the closure of a store derivation and specifying the
    option <option>--include-outputs</option>.</para>

    </listitem>

  </varlistentry>

  <varlistentry><term><option>--references</option></term>

    <listitem><para>Prints the set of <link linkend="gloss-reference">references</link> of the store paths
    <replaceable>paths</replaceable>, that is, their immediate
    dependencies.  (For <emphasis>all</emphasis> dependencies, use
    <option>--requisites</option>.)</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--referrers</option></term>

    <listitem><para>Prints the set of <emphasis>referrers</emphasis> of
    the store paths <replaceable>paths</replaceable>, that is, the
    store paths currently existing in the Nix store that refer to one
    of <replaceable>paths</replaceable>.  Note that contrary to the
    references, the set of referrers is not constant; it can change as
    store paths are added or removed.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--referrers-closure</option></term>

    <listitem><para>Prints the closure of the set of store paths
    <replaceable>paths</replaceable> under the referrers relation; that
    is, all store paths that directly or indirectly refer to one of
    <replaceable>paths</replaceable>.  These are all the path currently
    in the Nix store that are dependent on
    <replaceable>paths</replaceable>.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--deriver</option></term>
    <term><option>-d</option></term>

    <listitem><para>Prints the <link linkend="gloss-deriver">deriver</link> of the store paths
    <replaceable>paths</replaceable>.  If the path has no deriver
    (e.g., if it is a source file), or if the deriver is not known
    (e.g., in the case of a binary-only deployment), the string
    <literal>unknown-deriver</literal> is printed.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--graph</option></term>

    <listitem><para>Prints the references graph of the store paths
    <replaceable>paths</replaceable> in the format of the
    <command>dot</command> tool of AT&amp;T's <link xlink:href="http://www.graphviz.org/">Graphviz package</link>.
    This can be used to visualise dependency graphs.  To obtain a
    build-time dependency graph, apply this to a store derivation.  To
    obtain a runtime dependency graph, apply it to an output
    path.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--tree</option></term>

    <listitem><para>Prints the references graph of the store paths
    <replaceable>paths</replaceable> as a nested ASCII tree.
    References are ordered by descending closure size; this tends to
    flatten the tree, making it more readable.  The query only
    recurses into a store path when it is first encountered; this
    prevents a blowup of the tree representation of the
    graph.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--graphml</option></term>

    <listitem><para>Prints the references graph of the store paths
    <replaceable>paths</replaceable> in the <link xlink:href="http://graphml.graphdrawing.org/">GraphML</link> file format.
    This can be used to visualise dependency graphs. To obtain a
    build-time dependency graph, apply this to a store derivation. To
    obtain a runtime dependency graph, apply it to an output
    path.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--binding</option> <replaceable>name</replaceable></term>
    <term><option>-b</option> <replaceable>name</replaceable></term>

    <listitem><para>Prints the value of the attribute
    <replaceable>name</replaceable> (i.e., environment variable) of
    the store derivations <replaceable>paths</replaceable>.  It is an
    error for a derivation to not have the specified
    attribute.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--hash</option></term>

    <listitem><para>Prints the SHA-256 hash of the contents of the
    store paths <replaceable>paths</replaceable> (that is, the hash of
    the output of <command>nix-store --dump</command> on the given
    paths).  Since the hash is stored in the Nix database, this is a
    fast operation.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--size</option></term>

    <listitem><para>Prints the size in bytes of the contents of the
    store paths <replaceable>paths</replaceable> &#x2014; to be precise, the
    size of the output of <command>nix-store --dump</command> on the
    given paths.  Note that the actual disk space required by the
    store paths may be higher, especially on filesystems with large
    cluster sizes.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--roots</option></term>

    <listitem><para>Prints the garbage collector roots that point,
    directly or indirectly, at the store paths
    <replaceable>paths</replaceable>.</para></listitem>

  </varlistentry>

</variablelist>

</refsection>


<refsection><title>Examples</title>

<para>Print the closure (runtime dependencies) of the
<command>svn</command> program in the current user environment:

<screen>
$ nix-store -qR $(which svn)
/nix/store/5mbglq5ldqld8sj57273aljwkfvj22mc-subversion-1.1.4
/nix/store/9lz9yc6zgmc0vlqmn2ipcpkjlmbi51vv-glibc-2.3.4
<replaceable>...</replaceable></screen>

</para>

<para>Print the build-time dependencies of <command>svn</command>:

<screen>
$ nix-store -qR $(nix-store -qd $(which svn))
/nix/store/02iizgn86m42q905rddvg4ja975bk2i4-grep-2.5.1.tar.bz2.drv
/nix/store/07a2bzxmzwz5hp58nf03pahrv2ygwgs3-gcc-wrapper.sh
/nix/store/0ma7c9wsbaxahwwl04gbw3fcd806ski4-glibc-2.3.4.drv
<replaceable>... lots of other paths ...</replaceable></screen>

The difference with the previous example is that we ask the closure of
the derivation (<option>-qd</option>), not the closure of the output
path that contains <command>svn</command>.</para>

<para>Show the build-time dependencies as a tree:

<screen>
$ nix-store -q --tree $(nix-store -qd $(which svn))
/nix/store/7i5082kfb6yjbqdbiwdhhza0am2xvh6c-subversion-1.1.4.drv
+---/nix/store/d8afh10z72n8l1cr5w42366abiblgn54-builder.sh
+---/nix/store/fmzxmpjx2lh849ph0l36snfj9zdibw67-bash-3.0.drv
|   +---/nix/store/570hmhmx3v57605cqg9yfvvyh0nnb8k8-bash
|   +---/nix/store/p3srsbd8dx44v2pg6nbnszab5mcwx03v-builder.sh
<replaceable>...</replaceable></screen>

</para>

<para>Show all paths that depend on the same OpenSSL library as
<command>svn</command>:

<screen>
$ nix-store -q --referrers $(nix-store -q --binding openssl $(nix-store -qd $(which svn)))
/nix/store/23ny9l9wixx21632y2wi4p585qhva1q8-sylpheed-1.0.0
/nix/store/5mbglq5ldqld8sj57273aljwkfvj22mc-subversion-1.1.4
/nix/store/dpmvp969yhdqs7lm2r1a3gng7pyq6vy4-subversion-1.1.3
/nix/store/l51240xqsgg8a7yrbqdx1rfzyv6l26fx-lynx-2.8.5</screen>

</para>

<para>Show all paths that directly or indirectly depend on the Glibc
(C library) used by <command>svn</command>:

<screen>
$ nix-store -q --referrers-closure $(ldd $(which svn) | grep /libc.so | awk '{print $3}')
/nix/store/034a6h4vpz9kds5r6kzb9lhh81mscw43-libgnomeprintui-2.8.2
/nix/store/15l3yi0d45prm7a82pcrknxdh6nzmxza-gawk-3.1.4
<replaceable>...</replaceable></screen>

Note that <command>ldd</command> is a command that prints out the
dynamic libraries used by an ELF executable.</para>

<para>Make a picture of the runtime dependency graph of the current
user environment:

<screen>
$ nix-store -q --graph ~/.nix-profile | dot -Tps &gt; graph.ps
$ gv graph.ps</screen>

</para>

<para>Show every garbage collector root that points to a store path
that depends on <command>svn</command>:

<screen>
$ nix-store -q --roots $(which svn)
/nix/var/nix/profiles/default-81-link
/nix/var/nix/profiles/default-82-link
/nix/var/nix/profiles/per-user/eelco/profile-97-link
</screen>

</para>

</refsection>


</refsection>



<!--######################################################################-->

<!--
<refsection xml:id="rsec-nix-store-reg-val"><title>Operation <option>-XXX-register-validity</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-store</command>
  <arg choice='plain'><option>-XXX-register-validity</option></arg>
</cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>TODO</para>

</refsection>

</refsection>
-->



<!--######################################################################-->

<refsection><title>Operation <option>--add</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-store</command>
  <arg choice="plain"><option>--add</option></arg>
  <arg choice="plain" rep="repeat"><replaceable>paths</replaceable></arg>
</cmdsynopsis>

</refsection>

<refsection><title>Description</title>

<para>The operation <option>--add</option> adds the specified paths to
the Nix store.  It prints the resulting paths in the Nix store on
standard output.</para>

</refsection>

<refsection><title>Example</title>

<screen>
$ nix-store --add ./foo.c
/nix/store/m7lrha58ph6rcnv109yzx1nk1cj7k7zf-foo.c</screen>

</refsection>

</refsection>

<!--######################################################################-->

<refsection><title>Operation <option>--add-fixed</option></title>

<refsection><title>Synopsis</title>

<cmdsynopsis>
  <command>nix-store</command>
  <arg><option>--recursive</option></arg>
  <arg choice="plain"><option>--add-fixed</option></arg>
  <arg choice="plain"><replaceable>algorithm</replaceable></arg>
  <arg choice="plain" rep="repeat"><replaceable>paths</replaceable></arg>
</cmdsynopsis>

</refsection>

<refsection><title>Description</title>

<para>The operation <option>--add-fixed</option> adds the specified paths to
the Nix store.  Unlike <option>--add</option> paths are registered using the
specified hashing algorithm, resulting in the same output path as a fixed-output
derivation.  This can be used for sources that are not available from a public
url or broke since the download expression was written.
</para>

<para>This operation has the following options:

<variablelist>

  <varlistentry><term><option>--recursive</option></term>

    <listitem><para>
      Use recursive instead of flat hashing mode, used when adding directories
      to the store.
    </para></listitem>

  </varlistentry>

</variablelist>

</para>

</refsection>

<refsection><title>Example</title>

<screen>
$ nix-store --add-fixed sha256 ./hello-2.10.tar.gz
/nix/store/3x7dwzq014bblazs7kq20p9hyzz0qh8g-hello-2.10.tar.gz</screen>

</refsection>

</refsection>



<!--######################################################################-->

<refsection xml:id="refsec-nix-store-verify"><title>Operation <option>--verify</option></title>

<refsection>
  <title>Synopsis</title>
  <cmdsynopsis>
    <command>nix-store</command>
    <arg choice="plain"><option>--verify</option></arg>
    <arg><option>--check-contents</option></arg>
    <arg><option>--repair</option></arg>
  </cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>The operation <option>--verify</option> verifies the internal
consistency of the Nix database, and the consistency between the Nix
database and the Nix store.  Any inconsistencies encountered are
automatically repaired.  Inconsistencies are generally the result of
the Nix store or database being modified by non-Nix tools, or of bugs
in Nix itself.</para>

<para>This operation has the following options:

<variablelist>

  <varlistentry><term><option>--check-contents</option></term>

    <listitem><para>Checks that the contents of every valid store path
    has not been altered by computing a SHA-256 hash of the contents
    and comparing it with the hash stored in the Nix database at build
    time.  Paths that have been modified are printed out.  For large
    stores, <option>--check-contents</option> is obviously quite
    slow.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--repair</option></term>

    <listitem><para>If any valid path is missing from the store, or
    (if <option>--check-contents</option> is given) the contents of a
    valid path has been modified, then try to repair the path by
    redownloading it.  See <command>nix-store --repair-path</command>
    for details.</para></listitem>

  </varlistentry>

</variablelist>

</para>

</refsection>


</refsection>


<!--######################################################################-->

<refsection><title>Operation <option>--verify-path</option></title>

<refsection>
  <title>Synopsis</title>
  <cmdsynopsis>
    <command>nix-store</command>
    <arg choice="plain"><option>--verify-path</option></arg>
    <arg choice="plain" rep="repeat"><replaceable>paths</replaceable></arg>
  </cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>The operation <option>--verify-path</option> compares the
contents of the given store paths to their cryptographic hashes stored
in Nix&#x2019;s database.  For every changed path, it prints a warning
message.  The exit status is 0 if no path has changed, and 1
otherwise.</para>

</refsection>

<refsection><title>Example</title>

<para>To verify the integrity of the <command>svn</command> command and all its dependencies:

<screen>
$ nix-store --verify-path $(nix-store -qR $(which svn))
</screen>

</para>

</refsection>

</refsection>


<!--######################################################################-->

<refsection><title>Operation <option>--repair-path</option></title>

<refsection>
  <title>Synopsis</title>
  <cmdsynopsis>
    <command>nix-store</command>
    <arg choice="plain"><option>--repair-path</option></arg>
    <arg choice="plain" rep="repeat"><replaceable>paths</replaceable></arg>
  </cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>The operation <option>--repair-path</option> attempts to
&#x201C;repair&#x201D; the specified paths by redownloading them using the available
substituters.  If no substitutes are available, then repair is not
possible.</para>

<warning><para>During repair, there is a very small time window during
which the old path (if it exists) is moved out of the way and replaced
with the new path.  If repair is interrupted in between, then the
system may be left in a broken state (e.g., if the path contains a
critical system component like the GNU C Library).</para></warning>

</refsection>

<refsection><title>Example</title>

<screen>
$ nix-store --verify-path /nix/store/dj7a81wsm1ijwwpkks3725661h3263p5-glibc-2.13
path `/nix/store/dj7a81wsm1ijwwpkks3725661h3263p5-glibc-2.13' was modified!
  expected hash `2db57715ae90b7e31ff1f2ecb8c12ec1cc43da920efcbe3b22763f36a1861588',
  got `481c5aa5483ebc97c20457bb8bca24deea56550d3985cda0027f67fe54b808e4'

$ nix-store --repair-path /nix/store/dj7a81wsm1ijwwpkks3725661h3263p5-glibc-2.13
fetching path `/nix/store/d7a81wsm1ijwwpkks3725661h3263p5-glibc-2.13'...
&#x2026;
</screen>

</refsection>

</refsection>


<!--######################################################################-->

<refsection xml:id="refsec-nix-store-dump"><title>Operation <option>--dump</option></title>

<refsection>
  <title>Synopsis</title>
  <cmdsynopsis>
    <command>nix-store</command>
    <arg choice="plain"><option>--dump</option></arg>
    <arg choice="plain"><replaceable>path</replaceable></arg>
  </cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>The operation <option>--dump</option> produces a NAR (Nix
ARchive) file containing the contents of the file system tree rooted
at <replaceable>path</replaceable>.  The archive is written to
standard output.</para>

<para>A NAR archive is like a TAR or Zip archive, but it contains only
the information that Nix considers important.  For instance,
timestamps are elided because all files in the Nix store have their
timestamp set to 0 anyway.  Likewise, all permissions are left out
except for the execute bit, because all files in the Nix store have
444 or 555 permission.</para>

<para>Also, a NAR archive is <emphasis>canonical</emphasis>, meaning
that &#x201C;equal&#x201D; paths always produce the same NAR archive.  For instance,
directory entries are always sorted so that the actual on-disk order
doesn&#x2019;t influence the result.  This means that the cryptographic hash
of a NAR dump of a path is usable as a fingerprint of the contents of
the path.  Indeed, the hashes of store paths stored in Nix&#x2019;s database
(see <link linkend="refsec-nix-store-query"><literal>nix-store -q
--hash</literal></link>) are SHA-256 hashes of the NAR dump of each
store path.</para>

<para>NAR archives support filenames of unlimited length and 64-bit
file sizes.  They can contain regular files, directories, and symbolic
links, but not other types of files (such as device nodes).</para>

<para>A Nix archive can be unpacked using <literal>nix-store
--restore</literal>.</para>

</refsection>


</refsection>


<!--######################################################################-->

<refsection><title>Operation <option>--restore</option></title>

<refsection>
  <title>Synopsis</title>
  <cmdsynopsis>
    <command>nix-store</command>
    <arg choice="plain"><option>--restore</option></arg>
    <arg choice="plain"><replaceable>path</replaceable></arg>
  </cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>The operation <option>--restore</option> unpacks a NAR archive
to <replaceable>path</replaceable>, which must not already exist.  The
archive is read from standard input.</para>

</refsection>


</refsection>


<!--######################################################################-->

<refsection xml:id="refsec-nix-store-export"><title>Operation <option>--export</option></title>

<refsection>
  <title>Synopsis</title>
  <cmdsynopsis>
    <command>nix-store</command>
    <arg choice="plain"><option>--export</option></arg>
    <arg choice="plain" rep="repeat"><replaceable>paths</replaceable></arg>
  </cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>The operation <option>--export</option> writes a serialisation
of the specified store paths to standard output in a format that can
be imported into another Nix store with <command linkend="refsec-nix-store-import">nix-store --import</command>.  This
is like <command linkend="refsec-nix-store-dump">nix-store
--dump</command>, except that the NAR archive produced by that command
doesn&#x2019;t contain the necessary meta-information to allow it to be
imported into another Nix store (namely, the set of references of the
path).</para>

<para>This command does not produce a <emphasis>closure</emphasis> of
the specified paths, so if a store path references other store paths
that are missing in the target Nix store, the import will fail.  To
copy a whole closure, do something like:

<screen>
$ nix-store --export $(nix-store -qR <replaceable>paths</replaceable>) &gt; out</screen>

To import the whole closure again, run:

<screen>
$ nix-store --import &lt; out</screen>

</para>

</refsection>


</refsection>


<!--######################################################################-->

<refsection xml:id="refsec-nix-store-import"><title>Operation <option>--import</option></title>

<refsection>
  <title>Synopsis</title>
  <cmdsynopsis>
    <command>nix-store</command>
    <arg choice="plain"><option>--import</option></arg>
  </cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>The operation <option>--import</option> reads a serialisation of
a set of store paths produced by <command linkend="refsec-nix-store-export">nix-store --export</command> from
standard input and adds those store paths to the Nix store.  Paths
that already exist in the Nix store are ignored.  If a path refers to
another path that doesn&#x2019;t exist in the Nix store, the import
fails.</para>

</refsection>


</refsection>


<!--######################################################################-->

<refsection><title>Operation <option>--optimise</option></title>

<refsection>
  <title>Synopsis</title>
  <cmdsynopsis>
    <command>nix-store</command>
    <arg choice="plain"><option>--optimise</option></arg>
  </cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>The operation <option>--optimise</option> reduces Nix store disk
space usage by finding identical files in the store and hard-linking
them to each other.  It typically reduces the size of the store by
something like 25-35%.  Only regular files and symlinks are
hard-linked in this manner.  Files are considered identical when they
have the same NAR archive serialisation: that is, regular files must
have the same contents and permission (executable or non-executable),
and symlinks must have the same contents.</para>

<para>After completion, or when the command is interrupted, a report
on the achieved savings is printed on standard error.</para>

<para>Use <option>-vv</option> or <option>-vvv</option> to get some
progress indication.</para>

</refsection>

<refsection><title>Example</title>

<screen>
$ nix-store --optimise
hashing files in `/nix/store/qhqx7l2f1kmwihc9bnxs7rc159hsxnf3-gcc-4.1.1'
<replaceable>...</replaceable>
541838819 bytes (516.74 MiB) freed by hard-linking 54143 files;
there are 114486 files with equal contents out of 215894 files in total
</screen>

</refsection>


</refsection>


<!--######################################################################-->

<refsection><title>Operation <option>--read-log</option></title>

<refsection>
  <title>Synopsis</title>
  <cmdsynopsis>
    <command>nix-store</command>
    <group choice="req">
      <arg choice="plain"><option>--read-log</option></arg>
      <arg choice="plain"><option>-l</option></arg>
    </group>
    <arg choice="plain" rep="repeat"><replaceable>paths</replaceable></arg>
  </cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>The operation <option>--read-log</option> prints the build log
of the specified store paths on standard output.  The build log is
whatever the builder of a derivation wrote to standard output and
standard error.  If a store path is not a derivation, the deriver of
the store path is used.</para>

<para>Build logs are kept in
<filename>/nix/var/log/nix/drvs</filename>.  However, there is no
guarantee that a build log is available for any particular store path.
For instance, if the path was downloaded as a pre-built binary through
a substitute, then the log is unavailable.</para>

</refsection>

<refsection><title>Example</title>

<screen>
$ nix-store -l $(which ktorrent)
building /nix/store/dhc73pvzpnzxhdgpimsd9sw39di66ph1-ktorrent-2.2.1
unpacking sources
unpacking source archive /nix/store/p8n1jpqs27mgkjw07pb5269717nzf5f8-ktorrent-2.2.1.tar.gz
ktorrent-2.2.1/
ktorrent-2.2.1/NEWS
<replaceable>...</replaceable>
</screen>

</refsection>


</refsection>


<!--######################################################################-->

<refsection><title>Operation <option>--dump-db</option></title>

<refsection>
  <title>Synopsis</title>
  <cmdsynopsis>
    <command>nix-store</command>
    <arg choice="plain"><option>--dump-db</option></arg>
    <arg rep="repeat"><replaceable>paths</replaceable></arg>
  </cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>The operation <option>--dump-db</option> writes a dump of the
Nix database to standard output.  It can be loaded into an empty Nix
store using <option>--load-db</option>.  This is useful for making
backups and when migrating to different database schemas.</para>

<para>By default, <option>--dump-db</option> will dump the entire Nix
database.  When one or more store paths is passed, only the subset of
the Nix database for those store paths is dumped.  As with
<option>--export</option>, the user is responsible for passing all the
store paths for a closure.  See <option>--export</option> for an
example.</para>

</refsection>

</refsection>


<!--######################################################################-->

<refsection><title>Operation <option>--load-db</option></title>

<refsection>
  <title>Synopsis</title>
  <cmdsynopsis>
    <command>nix-store</command>
    <arg choice="plain"><option>--load-db</option></arg>
  </cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>The operation <option>--load-db</option> reads a dump of the Nix
database created by <option>--dump-db</option> from standard input and
loads it into the Nix database.</para>

</refsection>

</refsection>


<!--######################################################################-->

<refsection><title>Operation <option>--print-env</option></title>

<refsection>
  <title>Synopsis</title>
  <cmdsynopsis>
    <command>nix-store</command>
    <arg choice="plain"><option>--print-env</option></arg>
    <arg choice="plain"><replaceable>drvpath</replaceable></arg>
  </cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>The operation <option>--print-env</option> prints out the
environment of a derivation in a format that can be evaluated by a
shell.  The command line arguments of the builder are placed in the
variable <envar>_args</envar>.</para>

</refsection>

<refsection><title>Example</title>

<screen>
$ nix-store --print-env $(nix-instantiate '&lt;nixpkgs&gt;' -A firefox)
<replaceable>&#x2026;</replaceable>
export src; src='/nix/store/plpj7qrwcz94z2psh6fchsi7s8yihc7k-firefox-12.0.source.tar.bz2'
export stdenv; stdenv='/nix/store/7c8asx3yfrg5dg1gzhzyq2236zfgibnn-stdenv'
export system; system='x86_64-linux'
export _args; _args='-e /nix/store/9krlzvny65gdc8s7kpb6lkx8cd02c25c-default-builder.sh'
</screen>

</refsection>

</refsection>


<!--######################################################################-->

<refsection xml:id="rsec-nix-store-generate-binary-cache-key"><title>Operation <option>--generate-binary-cache-key</option></title>

<refsection>
  <title>Synopsis</title>
  <cmdsynopsis>
    <command>nix-store</command>
    <arg choice="plain">
      <option>--generate-binary-cache-key</option>
      <option>key-name</option>
      <option>secret-key-file</option>
      <option>public-key-file</option>
    </arg>
  </cmdsynopsis>
</refsection>

<refsection><title>Description</title>

<para>This command generates an <link xlink:href="http://ed25519.cr.yp.to/">Ed25519 key pair</link> that can
be used to create a signed binary cache. It takes three mandatory
parameters:

<orderedlist>

  <listitem><para>A key name, such as
  <literal>cache.example.org-1</literal>, that is used to look up keys
  on the client when it verifies signatures. It can be anything, but
  it&#x2019;s suggested to use the host name of your cache
  (e.g. <literal>cache.example.org</literal>) with a suffix denoting
  the number of the key (to be incremented every time you need to
  revoke a key).</para></listitem>

  <listitem><para>The file name where the secret key is to be
  stored.</para></listitem>

  <listitem><para>The file name where the public key is to be
  stored.</para></listitem>

</orderedlist>

</para>

</refsection>

</refsection>


<!--######################################################################-->

<refsection condition="manpage"><title>Environment variables</title>

<variablelist>
  <varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>IN_NIX_SHELL</envar></term>

  <listitem><para>Indicator that tells if the current environment was set up by
  <command>nix-shell</command>.  Since Nix 2.0 the values are
  <literal>"pure"</literal> and <literal>"impure"</literal></para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="env-NIX_PATH"><term><envar>NIX_PATH</envar></term>

  <listitem>

    <para>A colon-separated list of directories used to look up Nix
    expressions enclosed in angle brackets (i.e.,
    <literal>&lt;<replaceable>path</replaceable>&gt;</literal>).  For
    instance, the value

    <screen>
/home/eelco/Dev:/etc/nixos</screen>

    will cause Nix to look for paths relative to
    <filename>/home/eelco/Dev</filename> and
    <filename>/etc/nixos</filename>, in this order.  It is also
    possible to match paths against a prefix.  For example, the value

    <screen>
nixpkgs=/home/eelco/Dev/nixpkgs-branch:/etc/nixos</screen>

    will cause Nix to search for
    <literal>&lt;nixpkgs/<replaceable>path</replaceable>&gt;</literal> in
    <filename>/home/eelco/Dev/nixpkgs-branch/<replaceable>path</replaceable></filename>
    and
    <filename>/etc/nixos/nixpkgs/<replaceable>path</replaceable></filename>.</para>

    <para>If a path in the Nix search path starts with
    <literal>http://</literal> or <literal>https://</literal>, it is
    interpreted as the URL of a tarball that will be downloaded and
    unpacked to a temporary location. The tarball must consist of a
    single top-level directory. For example, setting
    <envar>NIX_PATH</envar> to

    <screen>
nixpkgs=https://github.com/NixOS/nixpkgs/archive/nixos-15.09.tar.gz</screen>

    tells Nix to download the latest revision in the Nixpkgs/NixOS
    15.09 channel.</para>

    <para>A following shorthand can be used to refer to the official channels:

    <screen>nixpkgs=channel:nixos-15.09</screen>
    </para>

    <para>The search path can be extended using the <option linkend="opt-I">-I</option> option, which takes precedence over
    <envar>NIX_PATH</envar>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_IGNORE_SYMLINK_STORE</envar></term>

  <listitem>

  <para>Normally, the Nix store directory (typically
  <filename>/nix/store</filename>) is not allowed to contain any
  symlink components.  This is to prevent &#x201C;impure&#x201D; builds.  Builders
  sometimes &#x201C;canonicalise&#x201D; paths by resolving all symlink components.
  Thus, builds on different machines (with
  <filename>/nix/store</filename> resolving to different locations)
  could yield different results.  This is generally not a problem,
  except when builds are deployed to machines where
  <filename>/nix/store</filename> resolves differently.  If you are
  sure that you&#x2019;re not going to do that, you can set
  <envar>NIX_IGNORE_SYMLINK_STORE</envar> to <envar>1</envar>.</para>

  <para>Note that if you&#x2019;re symlinking the Nix store so that you can
  put it on another file system than the root file system, on Linux
  you&#x2019;re better off using <literal>bind</literal> mount points, e.g.,

  <screen>
$ mkdir /nix
$ mount -o bind /mnt/otherdisk/nix /nix</screen>

  Consult the <citerefentry><refentrytitle>mount</refentrytitle>
  <manvolnum>8</manvolnum></citerefentry> manual page for details.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_STORE_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix store (default
  <filename><replaceable>prefix</replaceable>/store</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_DATA_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix static data
  directory (default
  <filename><replaceable>prefix</replaceable>/share</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_LOG_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix log directory
  (default <filename><replaceable>prefix</replaceable>/var/log/nix</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_STATE_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix state directory
  (default <filename><replaceable>prefix</replaceable>/var/nix</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_CONF_DIR</envar></term>

  <listitem><para>Overrides the location of the system Nix configuration
  directory (default
  <filename><replaceable>prefix</replaceable>/etc/nix</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_USER_CONF_FILES</envar></term>

  <listitem><para>Overrides the location of the user Nix configuration files
  to load from (defaults to the XDG spec locations). The variable is treated
  as a list separated by the <literal>:</literal> token.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>TMPDIR</envar></term>

  <listitem><para>Use the specified directory to store temporary
  files.  In particular, this includes temporary build directories;
  these can take up substantial amounts of disk space.  The default is
  <filename>/tmp</filename>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="envar-remote"><term><envar>NIX_REMOTE</envar></term>

  <listitem><para>This variable should be set to
  <literal>daemon</literal> if you want to use the Nix daemon to
  execute Nix operations. This is necessary in <link linkend="ssec-multi-user">multi-user Nix installations</link>.
  If the Nix daemon's Unix socket is at some non-standard path,
  this variable should be set to <literal>unix://path/to/socket</literal>.
  Otherwise, it should be left unset.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_SHOW_STATS</envar></term>

  <listitem><para>If set to <literal>1</literal>, Nix will print some
  evaluation statistics, such as the number of values
  allocated.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_COUNT_CALLS</envar></term>

  <listitem><para>If set to <literal>1</literal>, Nix will print how
  often functions were called during Nix expression evaluation.  This
  is useful for profiling your Nix expressions.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>GC_INITIAL_HEAP_SIZE</envar></term>

  <listitem><para>If Nix has been configured to use the Boehm garbage
  collector, this variable sets the initial size of the heap in bytes.
  It defaults to 384 MiB.  Setting it to a low value reduces memory
  consumption, but will increase runtime due to the overhead of
  garbage collection.</para></listitem>

</varlistentry>
</variablelist>

</refsection>


</refentry>

</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-utilities">

<title>Utilities</title>

<para>This section lists utilities that you can use when you
work with Nix.</para>

<refentry xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-nix-channel">

<refmeta>
  <refentrytitle>nix-channel</refentrytitle>
  <manvolnum>1</manvolnum>
  <refmiscinfo class="source">Nix</refmiscinfo>
  <refmiscinfo class="version">3.0</refmiscinfo>
</refmeta>

<refnamediv>
  <refname>nix-channel</refname>
  <refpurpose>manage Nix channels</refpurpose>
</refnamediv>

<refsynopsisdiv>
  <cmdsynopsis>
    <command>nix-channel</command>
    <group choice="req">
      <arg choice="plain"><option>--add</option> <replaceable>url</replaceable> <arg choice="opt"><replaceable>name</replaceable></arg></arg>
      <arg choice="plain"><option>--remove</option> <replaceable>name</replaceable></arg>
      <arg choice="plain"><option>--list</option></arg>
      <arg choice="plain"><option>--update</option> <arg rep="repeat"><replaceable>names</replaceable></arg></arg>
      <arg choice="plain"><option>--rollback</option> <arg choice="opt"><replaceable>generation</replaceable></arg></arg>
    </group>
  </cmdsynopsis>
</refsynopsisdiv>

<refsection><title>Description</title>

<para>A Nix channel is a mechanism that allows you to automatically
stay up-to-date with a set of pre-built Nix expressions.  A Nix
channel is just a URL that points to a place containing a set of Nix
expressions.  <phrase condition="manual">See also <xref linkend="sec-channels"/>.</phrase></para>
      
<para>To see the list of official NixOS channels, visit <link xlink:href="https://nixos.org/channels"/>.</para>

<para>This command has the following operations:

<variablelist>

  <varlistentry><term><option>--add</option> <replaceable>url</replaceable> [<replaceable>name</replaceable>]</term>

    <listitem><para>Adds a channel named
    <replaceable>name</replaceable> with URL
    <replaceable>url</replaceable> to the list of subscribed channels.
    If <replaceable>name</replaceable> is omitted, it defaults to the
    last component of <replaceable>url</replaceable>, with the
    suffixes <literal>-stable</literal> or
    <literal>-unstable</literal> removed.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--remove</option> <replaceable>name</replaceable></term>

    <listitem><para>Removes the channel named
    <replaceable>name</replaceable> from the list of subscribed
    channels.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--list</option></term>

    <listitem><para>Prints the names and URLs of all subscribed
    channels on standard output.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--update</option> [<replaceable>names</replaceable>&#x2026;]</term>

    <listitem><para>Downloads the Nix expressions of all subscribed
    channels (or only those included in
    <replaceable>names</replaceable> if specified) and makes them the
    default for <command>nix-env</command> operations (by symlinking
    them from the directory
    <filename>~/.nix-defexpr</filename>).</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--rollback</option> [<replaceable>generation</replaceable>]</term>

    <listitem><para>Reverts the previous call to <command>nix-channel
    --update</command>. Optionally, you can specify a specific channel
    generation number to restore.</para></listitem>

  </varlistentry>

</variablelist>

</para>

<para>Note that <option>--add</option> does not automatically perform
an update.</para>

<para>The list of subscribed channels is stored in
<filename>~/.nix-channels</filename>.</para>

</refsection>

<refsection><title>Examples</title>

<para>To subscribe to the Nixpkgs channel and install the GNU Hello package:</para>

<screen>
$ nix-channel --add https://nixos.org/channels/nixpkgs-unstable
$ nix-channel --update
$ nix-env -iA nixpkgs.hello</screen>

<para>You can revert channel updates using <option>--rollback</option>:</para>

<screen>
$ nix-instantiate --eval -E '(import &lt;nixpkgs&gt; {}).lib.version'
"14.04.527.0e935f1"

$ nix-channel --rollback
switching from generation 483 to 482

$ nix-instantiate --eval -E '(import &lt;nixpkgs&gt; {}).lib.version'
"14.04.526.dbadfad"
</screen>

</refsection>

<refsection><title>Files</title>

<variablelist>

  <varlistentry><term><filename>/nix/var/nix/profiles/per-user/<replaceable>username</replaceable>/channels</filename></term>

    <listitem><para><command>nix-channel</command> uses a
    <command>nix-env</command> profile to keep track of previous
    versions of the subscribed channels. Every time you run
    <command>nix-channel --update</command>, a new channel generation
    (that is, a symlink to the channel Nix expressions in the Nix store)
    is created. This enables <command>nix-channel --rollback</command>
    to revert to previous versions.</para></listitem>

  </varlistentry>

  <varlistentry><term><filename>~/.nix-defexpr/channels</filename></term>

    <listitem><para>This is a symlink to
    <filename>/nix/var/nix/profiles/per-user/<replaceable>username</replaceable>/channels</filename>. It
    ensures that <command>nix-env</command> can find your channels. In
    a multi-user installation, you may also have
    <filename>~/.nix-defexpr/channels_root</filename>, which links to
    the channels of the root user.</para></listitem>

  </varlistentry>

</variablelist>

</refsection>

<refsection><title>Channel format</title>

<para>A channel URL should point to a directory containing the
following files:</para>

<variablelist>

  <varlistentry><term><filename>nixexprs.tar.xz</filename></term>

    <listitem><para>A tarball containing Nix expressions and files
    referenced by them (such as build scripts and patches). At the
    top level, the tarball should contain a single directory. That
    directory must contain a file <filename>default.nix</filename>
    that serves as the channel&#x2019;s &#x201C;entry point&#x201D;.</para></listitem>

  </varlistentry>

</variablelist>

</refsection>

</refentry>
<refentry xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-nix-collect-garbage">
  
<refmeta>
  <refentrytitle>nix-collect-garbage</refentrytitle>
  <manvolnum>1</manvolnum>
  <refmiscinfo class="source">Nix</refmiscinfo>
  <refmiscinfo class="version">3.0</refmiscinfo>
</refmeta>

<refnamediv>
  <refname>nix-collect-garbage</refname>
  <refpurpose>delete unreachable store paths</refpurpose>
</refnamediv>

<refsynopsisdiv>
  <cmdsynopsis>
    <command>nix-collect-garbage</command>
    <arg><option>--delete-old</option></arg>
    <arg><option>-d</option></arg>
    <arg><option>--delete-older-than</option> <replaceable>period</replaceable></arg>
    <arg><option>--max-freed</option> <replaceable>bytes</replaceable></arg>
    <arg><option>--dry-run</option></arg>
  </cmdsynopsis>
</refsynopsisdiv>

<refsection><title>Description</title>

<para>The command <command>nix-collect-garbage</command> is mostly an
alias of <link linkend="rsec-nix-store-gc"><command>nix-store
--gc</command></link>, that is, it deletes all unreachable paths in
the Nix store to clean up your system.  However, it provides two
additional options: <option>-d</option> (<option>--delete-old</option>),
which deletes all old generations of all profiles in
<filename>/nix/var/nix/profiles</filename> by invoking
<literal>nix-env --delete-generations old</literal> on all profiles
(of course, this makes rollbacks to previous configurations
impossible); and
<option>--delete-older-than</option> <replaceable>period</replaceable>,
where period is a value such as <literal>30d</literal>, which deletes
all generations older than the specified number of days in all profiles
in <filename>/nix/var/nix/profiles</filename> (except for the generations
that were active at that point in time).
</para>

</refsection>

<refsection><title>Example</title>

<para>To delete from the Nix store everything that is not used by the
current generations of each profile, do

<screen>
$ nix-collect-garbage -d</screen>

</para>

</refsection>

</refentry>
<refentry xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" xml:id="sec-nix-copy-closure">

<refmeta>
  <refentrytitle>nix-copy-closure</refentrytitle>
  <manvolnum>1</manvolnum>
  <refmiscinfo class="source">Nix</refmiscinfo>
  <refmiscinfo class="version">3.0</refmiscinfo>
</refmeta>

<refnamediv>
  <refname>nix-copy-closure</refname>
  <refpurpose>copy a closure to or from a remote machine via SSH</refpurpose>
</refnamediv>

<refsynopsisdiv>
  <cmdsynopsis>
    <command>nix-copy-closure</command>
    <group>
      <arg choice="plain"><option>--to</option></arg>
      <arg choice="plain"><option>--from</option></arg>
    </group>
    <arg><option>--gzip</option></arg>
    <!--
    <arg><option>- -show-progress</option></arg>
    -->
    <arg><option>--include-outputs</option></arg>
    <group>
      <arg choice="plain"><option>--use-substitutes</option></arg>
      <arg choice="plain"><option>-s</option></arg>
    </group>
    <arg><option>-v</option></arg>
    <arg choice="plain">
      <replaceable>user@</replaceable><replaceable>machine</replaceable>
    </arg>
    <arg choice="plain"><replaceable>paths</replaceable></arg>
  </cmdsynopsis>
</refsynopsisdiv>


<refsection><title>Description</title>

<para><command>nix-copy-closure</command> gives you an easy and
efficient way to exchange software between machines.  Given one or
more Nix store <replaceable>paths</replaceable> on the local
machine, <command>nix-copy-closure</command> computes the closure of
those paths (i.e. all their dependencies in the Nix store), and copies
all paths in the closure to the remote machine via the
<command>ssh</command> (Secure Shell) command.  With the
<option>--from</option>, the direction is reversed:
the closure of <replaceable>paths</replaceable> on a remote machine is
copied to the Nix store on the local machine.</para>

<para>This command is efficient because it only sends the store paths
that are missing on the target machine.</para>

<para>Since <command>nix-copy-closure</command> calls
<command>ssh</command>, you may be asked to type in the appropriate
password or passphrase.  In fact, you may be asked
<emphasis>twice</emphasis> because <command>nix-copy-closure</command>
currently connects twice to the remote machine, first to get the set
of paths missing on the target machine, and second to send the dump of
those paths.  If this bothers you, use
<command>ssh-agent</command>.</para>


<refsection><title>Options</title>

<variablelist>

  <varlistentry><term><option>--to</option></term>

    <listitem><para>Copy the closure of
    <replaceable>paths</replaceable> from the local Nix store to the
    Nix store on <replaceable>machine</replaceable>.  This is the
    default.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--from</option></term>

    <listitem><para>Copy the closure of
    <replaceable>paths</replaceable> from the Nix store on
    <replaceable>machine</replaceable> to the local Nix
    store.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--gzip</option></term>

    <listitem><para>Enable compression of the SSH
    connection.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--include-outputs</option></term>

    <listitem><para>Also copy the outputs of store derivations
    included in the closure.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--use-substitutes</option> / <option>-s</option></term>

    <listitem><para>Attempt to download missing paths on the target
    machine using Nix&#x2019;s substitute mechanism.  Any paths that cannot
    be substituted on the target are still copied normally from the
    source.  This is useful, for instance, if the connection between
    the source and target machine is slow, but the connection between
    the target machine and <literal>nixos.org</literal> (the default
    binary cache server) is fast.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>-v</option></term>

    <listitem><para>Show verbose output.</para></listitem>

  </varlistentry>

</variablelist>

</refsection>


<refsection><title>Environment variables</title>

<variablelist>

  <varlistentry><term><envar>NIX_SSHOPTS</envar></term>

    <listitem><para>Additional options to be passed to
    <command>ssh</command> on the command line.</para></listitem>

  </varlistentry>

</variablelist>

</refsection>


<refsection><title>Examples</title>

<para>Copy Firefox with all its dependencies to a remote machine:

<screen>
$ nix-copy-closure --to alice@itchy.labs $(type -tP firefox)</screen>

</para>

<para>Copy Subversion from a remote machine and then install it into a
user environment:

<screen>
$ nix-copy-closure --from alice@itchy.labs \
    /nix/store/0dj0503hjxy5mbwlafv1rsbdiyx1gkdy-subversion-1.4.4
$ nix-env -i /nix/store/0dj0503hjxy5mbwlafv1rsbdiyx1gkdy-subversion-1.4.4
</screen>

</para>

</refsection>


</refsection>

</refentry>
<refentry xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-nix-daemon">

<refmeta>
  <refentrytitle>nix-daemon</refentrytitle>
  <manvolnum>8</manvolnum>
  <refmiscinfo class="source">Nix</refmiscinfo>
  <refmiscinfo class="version">3.0</refmiscinfo>
</refmeta>

<refnamediv>
  <refname>nix-daemon</refname>
  <refpurpose>Nix multi-user support daemon</refpurpose>
</refnamediv>

<refsynopsisdiv>
  <cmdsynopsis>
    <command>nix-daemon</command>
  </cmdsynopsis>
</refsynopsisdiv>


<refsection><title>Description</title>

<para>The Nix daemon is necessary in multi-user Nix installations.  It
performs build actions and other operations on the Nix store on behalf
of unprivileged users.</para>


</refsection>

</refentry>
<refentry xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-nix-hash">
  
<refmeta>
  <refentrytitle>nix-hash</refentrytitle>
  <manvolnum>1</manvolnum>
  <refmiscinfo class="source">Nix</refmiscinfo>
  <refmiscinfo class="version">3.0</refmiscinfo>
</refmeta>

<refnamediv>
  <refname>nix-hash</refname>
  <refpurpose>compute the cryptographic hash of a path</refpurpose>
</refnamediv>

<refsynopsisdiv>
  <cmdsynopsis>
    <command>nix-hash</command>
    <arg><option>--flat</option></arg>
    <arg><option>--base32</option></arg>
    <arg><option>--truncate</option></arg>
    <arg><option>--type</option> <replaceable>hashAlgo</replaceable></arg>
    <arg choice="plain" rep="repeat"><replaceable>path</replaceable></arg>
  </cmdsynopsis>
  <cmdsynopsis>
    <command>nix-hash</command>
    <arg choice="plain"><option>--to-base16</option></arg>
    <arg choice="plain" rep="repeat"><replaceable>hash</replaceable></arg>
  </cmdsynopsis>
  <cmdsynopsis>
    <command>nix-hash</command>
    <arg choice="plain"><option>--to-base32</option></arg>
    <arg choice="plain" rep="repeat"><replaceable>hash</replaceable></arg>
  </cmdsynopsis>
</refsynopsisdiv>


<refsection><title>Description</title>

<para>The command <command>nix-hash</command> computes the
cryptographic hash of the contents of each
<replaceable>path</replaceable> and prints it on standard output.  By
default, it computes an MD5 hash, but other hash algorithms are
available as well.  The hash is printed in hexadecimal.  To generate
the same hash as <command>nix-prefetch-url</command> you have to
specify multiple arguments, see below for an example.</para>

<para>The hash is computed over a <emphasis>serialisation</emphasis>
of each path: a dump of the file system tree rooted at the path.  This
allows directories and symlinks to be hashed as well as regular files.
The dump is in the <emphasis>NAR format</emphasis> produced by <link linkend="refsec-nix-store-dump"><command>nix-store</command>
<option>--dump</option></link>.  Thus, <literal>nix-hash
<replaceable>path</replaceable></literal> yields the same
cryptographic hash as <literal>nix-store --dump
<replaceable>path</replaceable> | md5sum</literal>.</para>

</refsection>


<refsection><title>Options</title>

<variablelist>
  
  <varlistentry><term><option>--flat</option></term>

    <listitem><para>Print the cryptographic hash of the contents of
    each regular file <replaceable>path</replaceable>.  That is, do
    not compute the hash over the dump of
    <replaceable>path</replaceable>.  The result is identical to that
    produced by the GNU commands <command>md5sum</command> and
    <command>sha1sum</command>.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--base32</option></term>

    <listitem><para>Print the hash in a base-32 representation rather
    than hexadecimal.  This base-32 representation is more compact and
    can be used in Nix expressions (such as in calls to
    <function>fetchurl</function>).</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--truncate</option></term>

    <listitem><para>Truncate hashes longer than 160 bits (such as
    SHA-256) to 160 bits.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--type</option> <replaceable>hashAlgo</replaceable></term>

    <listitem><para>Use the specified cryptographic hash algorithm,
    which can be one of <literal>md5</literal>,
    <literal>sha1</literal>, and
    <literal>sha256</literal>.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--to-base16</option></term>

    <listitem><para>Don&#x2019;t hash anything, but convert the base-32 hash
    representation <replaceable>hash</replaceable> to
    hexadecimal.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--to-base32</option></term>

    <listitem><para>Don&#x2019;t hash anything, but convert the hexadecimal
    hash representation <replaceable>hash</replaceable> to
    base-32.</para></listitem>

  </varlistentry>

</variablelist>

</refsection>


<refsection><title>Examples</title>

<para>Computing the same hash as <command>nix-prefetch-url</command>:
<screen>
$ nix-prefetch-url file://&lt;(echo test)
1lkgqb6fclns49861dwk9rzb6xnfkxbpws74mxnx01z9qyv1pjpj
$ nix-hash --type sha256 --flat --base32 &lt;(echo test)
1lkgqb6fclns49861dwk9rzb6xnfkxbpws74mxnx01z9qyv1pjpj
</screen>
</para>

<para>Computing hashes:

<screen>
$ mkdir test
$ echo "hello" &gt; test/world

$ nix-hash test/ <lineannotation>(MD5 hash; default)</lineannotation>
8179d3caeff1869b5ba1744e5a245c04

$ nix-store --dump test/ | md5sum <lineannotation>(for comparison)</lineannotation>
8179d3caeff1869b5ba1744e5a245c04  -

$ nix-hash --type sha1 test/
e4fd8ba5f7bbeaea5ace89fe10255536cd60dab6

$ nix-hash --type sha1 --base32 test/
nvd61k9nalji1zl9rrdfmsmvyyjqpzg4

$ nix-hash --type sha256 --flat test/
error: reading file `test/': Is a directory

$ nix-hash --type sha256 --flat test/world
5891b5b522d5df086d0ff0b110fbd9d21bb4fc7163af34d08286a2e846f6be03</screen>

</para>

<para>Converting between hexadecimal and base-32:

<screen>
$ nix-hash --type sha1 --to-base32 e4fd8ba5f7bbeaea5ace89fe10255536cd60dab6
nvd61k9nalji1zl9rrdfmsmvyyjqpzg4

$ nix-hash --type sha1 --to-base16 nvd61k9nalji1zl9rrdfmsmvyyjqpzg4
e4fd8ba5f7bbeaea5ace89fe10255536cd60dab6</screen>

</para>

</refsection>


</refentry>
<refentry xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-nix-instantiate">

<refmeta>
  <refentrytitle>nix-instantiate</refentrytitle>
  <manvolnum>1</manvolnum>
  <refmiscinfo class="source">Nix</refmiscinfo>
  <refmiscinfo class="version">3.0</refmiscinfo>
</refmeta>

<refnamediv>
  <refname>nix-instantiate</refname>
  <refpurpose>instantiate store derivations from Nix expressions</refpurpose>
</refnamediv>

<refsynopsisdiv>
  <cmdsynopsis>
    <command>nix-instantiate</command>
    <group>
      <arg choice="plain"><option>--parse</option></arg>
      <arg choice="plain">
        <option>--eval</option>
        <arg><option>--strict</option></arg>
        <arg><option>--json</option></arg>
        <arg><option>--xml</option></arg>
      </arg>
    </group>
    <arg><option>--read-write-mode</option></arg>
    <arg><option>--arg</option> <replaceable>name</replaceable> <replaceable>value</replaceable></arg>
    <arg>
      <group choice="req">
        <arg choice="plain"><option>--attr</option></arg>
        <arg choice="plain"><option>-A</option></arg>
      </group>
      <replaceable>attrPath</replaceable>
    </arg>
    <arg><option>--add-root</option> <replaceable>path</replaceable></arg>
    <arg><option>--indirect</option></arg>
    <group>
      <arg choice="plain"><option>--expr</option></arg>
      <arg choice="plain"><option>-E</option></arg>
    </group>
    <arg choice="plain" rep="repeat"><replaceable>files</replaceable></arg>
  </cmdsynopsis>
  <cmdsynopsis>
    <command>nix-instantiate</command>
    <arg choice="plain"><option>--find-file</option></arg>
    <arg choice="plain" rep="repeat"><replaceable>files</replaceable></arg>
  </cmdsynopsis>
</refsynopsisdiv>


<refsection><title>Description</title>

<para>The command <command>nix-instantiate</command> generates <link linkend="gloss-derivation">store derivations</link> from (high-level)
Nix expressions.  It evaluates the Nix expressions in each of
<replaceable>files</replaceable> (which defaults to
<replaceable>./default.nix</replaceable>).  Each top-level expression
should evaluate to a derivation, a list of derivations, or a set of
derivations.  The paths of the resulting store derivations are printed
on standard output.</para>

<para>If <replaceable>files</replaceable> is the character
<literal>-</literal>, then a Nix expression will be read from standard
input.</para>

<para condition="manual">See also <xref linkend="sec-common-options"/> for a list of common options.</para>

</refsection>


<refsection><title>Options</title>

<variablelist>

  <varlistentry>
    <term><option>--add-root</option> <replaceable>path</replaceable></term>
    <term><option>--indirect</option></term>

    <listitem><para>See the <link linkend="opt-add-root">corresponding
    options</link> in <command>nix-store</command>.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--parse</option></term>

    <listitem><para>Just parse the input files, and print their
    abstract syntax trees on standard output in ATerm
    format.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--eval</option></term>

    <listitem><para>Just parse and evaluate the input files, and print
    the resulting values on standard output.  No instantiation of
    store derivations takes place.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--find-file</option></term>

    <listitem><para>Look up the given files in Nix&#x2019;s search path (as
    specified by the <envar linkend="env-NIX_PATH">NIX_PATH</envar>
    environment variable).  If found, print the corresponding absolute
    paths on standard output.  For instance, if
    <envar>NIX_PATH</envar> is
    <literal>nixpkgs=/home/alice/nixpkgs</literal>, then
    <literal>nix-instantiate --find-file nixpkgs/default.nix</literal>
    will print
    <literal>/home/alice/nixpkgs/default.nix</literal>.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--strict</option></term>

    <listitem><para>When used with <option>--eval</option>,
    recursively evaluate list elements and attributes.  Normally, such
    sub-expressions are left unevaluated (since the Nix expression
    language is lazy).</para>

    <warning><para>This option can cause non-termination, because lazy
    data structures can be infinitely large.</para></warning>

    </listitem>

  </varlistentry>

  <varlistentry><term><option>--json</option></term>

    <listitem><para>When used with <option>--eval</option>, print the resulting
    value as an JSON representation of the abstract syntax tree rather
    than as an ATerm.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--xml</option></term>

    <listitem><para>When used with <option>--eval</option>, print the resulting
    value as an XML representation of the abstract syntax tree rather than as
    an ATerm. The schema is the same as that used by the <link linkend="builtin-toXML"><function>toXML</function> built-in</link>.
    </para></listitem>

  </varlistentry>

  <varlistentry><term><option>--read-write-mode</option></term>

    <listitem><para>When used with <option>--eval</option>, perform
    evaluation in read/write mode so nix language features that
    require it will still work (at the cost of needing to do
    instantiation of every evaluated derivation). If this option is
    not enabled, there may be uninstantiated store paths in the final
    output.</para>

    </listitem>

  </varlistentry>

</variablelist>

<variablelist condition="manpage">
  <varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--help</option></term>

  <listitem><para>Prints out a summary of the command syntax and
  exits.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--version</option></term>

  <listitem><para>Prints out the Nix version number on standard output
  and exits.</para></listitem>
</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--verbose</option> / <option>-v</option></term>

  <listitem>

  <para>Increases the level of verbosity of diagnostic messages
  printed on standard error.  For each Nix operation, the information
  printed on standard output is well-defined; any diagnostic
  information is printed on standard error, never on standard
  output.</para>

  <para>This option may be specified repeatedly.  Currently, the
  following verbosity levels exist:</para>

  <variablelist>

    <varlistentry><term>0</term>
    <listitem><para>&#x201C;Errors only&#x201D;: only print messages
    explaining why the Nix invocation failed.</para></listitem>
    </varlistentry>

    <varlistentry><term>1</term>
    <listitem><para>&#x201C;Informational&#x201D;: print
    <emphasis>useful</emphasis> messages about what Nix is doing.
    This is the default.</para></listitem>
    </varlistentry>

    <varlistentry><term>2</term>
    <listitem><para>&#x201C;Talkative&#x201D;: print more informational
    messages.</para></listitem>
    </varlistentry>

    <varlistentry><term>3</term>
    <listitem><para>&#x201C;Chatty&#x201D;: print even more
    informational messages.</para></listitem>
    </varlistentry>

    <varlistentry><term>4</term>
    <listitem><para>&#x201C;Debug&#x201D;: print debug
    information.</para></listitem>
    </varlistentry>

    <varlistentry><term>5</term>
    <listitem><para>&#x201C;Vomit&#x201D;: print vast amounts of debug
    information.</para></listitem>
    </varlistentry>

  </variablelist>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--quiet</option></term>

  <listitem>

  <para>Decreases the level of verbosity of diagnostic messages
  printed on standard error.  This is the inverse option to
  <option>-v</option> / <option>--verbose</option>.
  </para>

  <para>This option may be specified repeatedly.  See the previous
  verbosity levels list.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-log-format"><term><option>--log-format</option> <replaceable>format</replaceable></term>

  <listitem>

  <para>This option can be used to change the output of the log format, with
  <replaceable>format</replaceable> being one of:</para>

  <variablelist>

    <varlistentry><term>raw</term>
    <listitem><para>This is the raw format, as outputted by nix-build.</para></listitem>
    </varlistentry>

    <varlistentry><term>internal-json</term>
    <listitem><para>Outputs the logs in a structured manner. NOTE: the json schema is not guarantees to be stable between releases.</para></listitem>
    </varlistentry>

    <varlistentry><term>bar</term>
    <listitem><para>Only display a progress bar during the builds.</para></listitem>
    </varlistentry>

    <varlistentry><term>bar-with-logs</term>
    <listitem><para>Display the raw logs, with the progress bar at the bottom.</para></listitem>
    </varlistentry>

  </variablelist>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--no-build-output</option> / <option>-Q</option></term>

  <listitem><para>By default, output written by builders to standard
  output and standard error is echoed to the Nix command's standard
  error.  This option suppresses this behaviour.  Note that the
  builder's standard output and error are always written to a log file
  in
  <filename><replaceable>prefix</replaceable>/nix/var/log/nix</filename>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-max-jobs"><term><option>--max-jobs</option> / <option>-j</option>
<replaceable>number</replaceable></term>

  <listitem>

  <para>Sets the maximum number of build jobs that Nix will
  perform in parallel to the specified number.  Specify
  <literal>auto</literal> to use the number of CPUs in the system.
  The default is specified by the <link linkend="conf-max-jobs"><literal>max-jobs</literal></link>
  configuration setting, which itself defaults to
  <literal>1</literal>.  A higher value is useful on SMP systems or to
  exploit I/O latency.</para>

  <para> Setting it to <literal>0</literal> disallows building on the local
  machine, which is useful when you want builds to happen only on remote
  builders.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-cores"><term><option>--cores</option></term>

  <listitem><para>Sets the value of the <envar>NIX_BUILD_CORES</envar>
  environment variable in the invocation of builders.  Builders can
  use this variable at their discretion to control the maximum amount
  of parallelism.  For instance, in Nixpkgs, if the derivation
  attribute <varname>enableParallelBuilding</varname> is set to
  <literal>true</literal>, the builder passes the
  <option>-j<replaceable>N</replaceable></option> flag to GNU Make.
  It defaults to the value of the <link linkend="conf-cores"><literal>cores</literal></link>
  configuration setting, if set, or <literal>1</literal> otherwise.
  The value <literal>0</literal> means that the builder should use all
  available CPU cores in the system.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-max-silent-time"><term><option>--max-silent-time</option></term>

  <listitem><para>Sets the maximum number of seconds that a builder
  can go without producing any data on standard output or standard
  error.  The default is specified by the <link linkend="conf-max-silent-time"><literal>max-silent-time</literal></link>
  configuration setting.  <literal>0</literal> means no
  time-out.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-timeout"><term><option>--timeout</option></term>

  <listitem><para>Sets the maximum number of seconds that a builder
  can run.  The default is specified by the <link linkend="conf-timeout"><literal>timeout</literal></link>
  configuration setting.  <literal>0</literal> means no
  timeout.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--keep-going</option> / <option>-k</option></term>

  <listitem><para>Keep going in case of failed builds, to the
  greatest extent possible.  That is, if building an input of some
  derivation fails, Nix will still build the other inputs, but not the
  derivation itself.  Without this option, Nix stops if any build
  fails (except for builds of substitutes), possibly killing builds in
  progress (in case of parallel or distributed builds).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--keep-failed</option> / <option>-K</option></term>

  <listitem><para>Specifies that in case of a build failure, the
  temporary directory (usually in <filename>/tmp</filename>) in which
  the build takes place should not be deleted.  The path of the build
  directory is printed as an informational message.
    </para>
  </listitem>
</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--fallback</option></term>

  <listitem>

  <para>Whenever Nix attempts to build a derivation for which
  substitutes are known for each output path, but realising the output
  paths through the substitutes fails, fall back on building the
  derivation.</para>

  <para>The most common scenario in which this is useful is when we
  have registered substitutes in order to perform binary distribution
  from, say, a network repository.  If the repository is down, the
  realisation of the derivation will fail.  When this option is
  specified, Nix will build the derivation instead.  Thus,
  installation from binaries falls back on installation from source.
  This option is not the default since it is generally not desirable
  for a transient failure in obtaining the substitutes to lead to a
  full build from source (with the related consumption of
  resources).</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--no-build-hook</option></term>

  <listitem>

  <para>Disables the build hook mechanism.  This allows to ignore remote
  builders if they are setup on the machine.</para>

  <para>It's useful in cases where the bandwidth between the client and the
  remote builder is too low.  In that case it can take more time to upload the
  sources to the remote builder and fetch back the result than to do the
  computation locally.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--readonly-mode</option></term>

  <listitem><para>When this option is used, no attempt is made to open
  the Nix database.  Most Nix operations do need database access, so
  those operations will fail.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--arg</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>This option is accepted by
  <command>nix-env</command>, <command>nix-instantiate</command>,
  <command>nix-shell</command> and <command>nix-build</command>.
  When evaluating Nix expressions, the expression evaluator will
  automatically try to call functions that
  it encounters.  It can automatically call functions for which every
  argument has a <link linkend="ss-functions">default value</link>
  (e.g., <literal>{ <replaceable>argName</replaceable> ?
  <replaceable>defaultValue</replaceable> }:
  <replaceable>...</replaceable></literal>).  With
  <option>--arg</option>, you can also call functions that have
  arguments without a default value (or override a default value).
  That is, if the evaluator encounters a function with an argument
  named <replaceable>name</replaceable>, it will call it with value
  <replaceable>value</replaceable>.</para>

  <para>For instance, the top-level <literal>default.nix</literal> in
  Nixpkgs is actually a function:

<programlisting>
{ # The system (e.g., `i686-linux') for which to build the packages.
  system ? builtins.currentSystem
  <replaceable>...</replaceable>
}: <replaceable>...</replaceable></programlisting>

  So if you call this Nix expression (e.g., when you do
  <literal>nix-env -i <replaceable>pkgname</replaceable></literal>),
  the function will be called automatically using the value <link linkend="builtin-currentSystem"><literal>builtins.currentSystem</literal></link>
  for the <literal>system</literal> argument.  You can override this
  using <option>--arg</option>, e.g., <literal>nix-env -i
  <replaceable>pkgname</replaceable> --arg system
  \"i686-freebsd\"</literal>.  (Note that since the argument is a Nix
  string literal, you have to escape the quotes.)</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--argstr</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>This option is like <option>--arg</option>, only the
  value is not a Nix expression but a string.  So instead of
  <literal>--arg system \"i686-linux\"</literal> (the outer quotes are
  to keep the shell happy) you can say <literal>--argstr system
  i686-linux</literal>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-attr"><term><option>--attr</option> / <option>-A</option>
<replaceable>attrPath</replaceable></term>

  <listitem><para>Select an attribute from the top-level Nix
  expression being evaluated.  (<command>nix-env</command>,
  <command>nix-instantiate</command>, <command>nix-build</command> and
  <command>nix-shell</command> only.)  The <emphasis>attribute
  path</emphasis> <replaceable>attrPath</replaceable> is a sequence of
  attribute names separated by dots.  For instance, given a top-level
  Nix expression <replaceable>e</replaceable>, the attribute path
  <literal>xorg.xorgserver</literal> would cause the expression
  <literal><replaceable>e</replaceable>.xorg.xorgserver</literal> to
  be used.  See <link linkend="refsec-nix-env-install-examples"><command>nix-env
  --install</command></link> for some concrete examples.</para>

  <para>In addition to attribute names, you can also specify array
  indices.  For instance, the attribute path
  <literal>foo.3.bar</literal> selects the <literal>bar</literal>
  attribute of the fourth element of the array in the
  <literal>foo</literal> attribute of the top-level
  expression.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--expr</option> / <option>-E</option></term>

  <listitem><para>Interpret the command line arguments as a list of
  Nix expressions to be parsed and evaluated, rather than as a list
  of file names of Nix expressions.
  (<command>nix-instantiate</command>, <command>nix-build</command>
  and <command>nix-shell</command> only.)</para>

  <para>For <command>nix-shell</command>, this option is commonly used
  to give you a shell in which you can build the packages returned
  by the expression. If you want to get a shell which contain the
  <emphasis>built</emphasis> packages ready for use, give your
  expression to the <command>nix-shell -p</command> convenience flag
  instead.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="opt-I"><term><option>-I</option> <replaceable>path</replaceable></term>

  <listitem><para>Add a path to the Nix expression search path.  This
  option may be given multiple times.  See the <envar linkend="env-NIX_PATH">NIX_PATH</envar> environment variable for
  information on the semantics of the Nix search path.  Paths added
  through <option>-I</option> take precedence over
  <envar>NIX_PATH</envar>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--option</option> <replaceable>name</replaceable> <replaceable>value</replaceable></term>

  <listitem><para>Set the Nix configuration option
  <replaceable>name</replaceable> to <replaceable>value</replaceable>.
  This overrides settings in the Nix configuration file (see
  <citerefentry><refentrytitle>nix.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><option>--repair</option></term>

  <listitem><para>Fix corrupted or missing store paths by
  redownloading or rebuilding them.  Note that this is slow because it
  requires computing a cryptographic hash of the contents of every
  path in the closure of the build.  Also note the warning under
  <command>nix-store --repair-path</command>.</para></listitem>

</varlistentry>
</variablelist>

</refsection>


<refsection><title>Examples</title>

<para>Instantiating store derivations from a Nix expression, and
building them using <command>nix-store</command>:

<screen>
$ nix-instantiate test.nix <lineannotation>(instantiate)</lineannotation>
/nix/store/cigxbmvy6dzix98dxxh9b6shg7ar5bvs-perl-BerkeleyDB-0.26.drv

$ nix-store -r $(nix-instantiate test.nix) <lineannotation>(build)</lineannotation>
<replaceable>...</replaceable>
/nix/store/qhqk4n8ci095g3sdp93x7rgwyh9rdvgk-perl-BerkeleyDB-0.26 <lineannotation>(output path)</lineannotation>

$ ls -l /nix/store/qhqk4n8ci095g3sdp93x7rgwyh9rdvgk-perl-BerkeleyDB-0.26
dr-xr-xr-x    2 eelco    users        4096 1970-01-01 01:00 lib
...</screen>

</para>

<para>You can also give a Nix expression on the command line:

<screen>
$ nix-instantiate -E 'with import &lt;nixpkgs&gt; { }; hello'
/nix/store/j8s4zyv75a724q38cb0r87rlczaiag4y-hello-2.8.drv
</screen>

This is equivalent to:

<screen>
$ nix-instantiate '&lt;nixpkgs&gt;' -A hello
</screen>

</para>

<para>Parsing and evaluating Nix expressions:

<screen>
$ nix-instantiate --parse -E '1 + 2'
1 + 2

$ nix-instantiate --eval -E '1 + 2'
3

$ nix-instantiate --eval --xml -E '1 + 2'
<![CDATA[<?xml version='1.0' encoding='utf-8'?>
<expr>
  <int value="3" />
</expr>]]></screen>

</para>

<para>The difference between non-strict and strict evaluation:

<screen>
$ nix-instantiate --eval --xml -E 'rec { x = "foo"; y = x; }'
<replaceable>...</replaceable><![CDATA[
  <attr name="x">
    <string value="foo" />
  </attr>
  <attr name="y">
    <unevaluated />
  </attr>]]>
<replaceable>...</replaceable></screen>

Note that <varname>y</varname> is left unevaluated (the XML
representation doesn&#x2019;t attempt to show non-normal forms).

<screen>
$ nix-instantiate --eval --xml --strict -E 'rec { x = "foo"; y = x; }'
<replaceable>...</replaceable><![CDATA[
  <attr name="x">
    <string value="foo" />
  </attr>
  <attr name="y">
    <string value="foo" />
  </attr>]]>
<replaceable>...</replaceable></screen>

</para>

</refsection>


<refsection condition="manpage"><title>Environment variables</title>

<variablelist>
  <varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>IN_NIX_SHELL</envar></term>

  <listitem><para>Indicator that tells if the current environment was set up by
  <command>nix-shell</command>.  Since Nix 2.0 the values are
  <literal>"pure"</literal> and <literal>"impure"</literal></para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="env-NIX_PATH"><term><envar>NIX_PATH</envar></term>

  <listitem>

    <para>A colon-separated list of directories used to look up Nix
    expressions enclosed in angle brackets (i.e.,
    <literal>&lt;<replaceable>path</replaceable>&gt;</literal>).  For
    instance, the value

    <screen>
/home/eelco/Dev:/etc/nixos</screen>

    will cause Nix to look for paths relative to
    <filename>/home/eelco/Dev</filename> and
    <filename>/etc/nixos</filename>, in this order.  It is also
    possible to match paths against a prefix.  For example, the value

    <screen>
nixpkgs=/home/eelco/Dev/nixpkgs-branch:/etc/nixos</screen>

    will cause Nix to search for
    <literal>&lt;nixpkgs/<replaceable>path</replaceable>&gt;</literal> in
    <filename>/home/eelco/Dev/nixpkgs-branch/<replaceable>path</replaceable></filename>
    and
    <filename>/etc/nixos/nixpkgs/<replaceable>path</replaceable></filename>.</para>

    <para>If a path in the Nix search path starts with
    <literal>http://</literal> or <literal>https://</literal>, it is
    interpreted as the URL of a tarball that will be downloaded and
    unpacked to a temporary location. The tarball must consist of a
    single top-level directory. For example, setting
    <envar>NIX_PATH</envar> to

    <screen>
nixpkgs=https://github.com/NixOS/nixpkgs/archive/nixos-15.09.tar.gz</screen>

    tells Nix to download the latest revision in the Nixpkgs/NixOS
    15.09 channel.</para>

    <para>A following shorthand can be used to refer to the official channels:

    <screen>nixpkgs=channel:nixos-15.09</screen>
    </para>

    <para>The search path can be extended using the <option linkend="opt-I">-I</option> option, which takes precedence over
    <envar>NIX_PATH</envar>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_IGNORE_SYMLINK_STORE</envar></term>

  <listitem>

  <para>Normally, the Nix store directory (typically
  <filename>/nix/store</filename>) is not allowed to contain any
  symlink components.  This is to prevent &#x201C;impure&#x201D; builds.  Builders
  sometimes &#x201C;canonicalise&#x201D; paths by resolving all symlink components.
  Thus, builds on different machines (with
  <filename>/nix/store</filename> resolving to different locations)
  could yield different results.  This is generally not a problem,
  except when builds are deployed to machines where
  <filename>/nix/store</filename> resolves differently.  If you are
  sure that you&#x2019;re not going to do that, you can set
  <envar>NIX_IGNORE_SYMLINK_STORE</envar> to <envar>1</envar>.</para>

  <para>Note that if you&#x2019;re symlinking the Nix store so that you can
  put it on another file system than the root file system, on Linux
  you&#x2019;re better off using <literal>bind</literal> mount points, e.g.,

  <screen>
$ mkdir /nix
$ mount -o bind /mnt/otherdisk/nix /nix</screen>

  Consult the <citerefentry><refentrytitle>mount</refentrytitle>
  <manvolnum>8</manvolnum></citerefentry> manual page for details.</para>

  </listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_STORE_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix store (default
  <filename><replaceable>prefix</replaceable>/store</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_DATA_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix static data
  directory (default
  <filename><replaceable>prefix</replaceable>/share</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_LOG_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix log directory
  (default <filename><replaceable>prefix</replaceable>/var/log/nix</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_STATE_DIR</envar></term>

  <listitem><para>Overrides the location of the Nix state directory
  (default <filename><replaceable>prefix</replaceable>/var/nix</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_CONF_DIR</envar></term>

  <listitem><para>Overrides the location of the system Nix configuration
  directory (default
  <filename><replaceable>prefix</replaceable>/etc/nix</filename>).</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_USER_CONF_FILES</envar></term>

  <listitem><para>Overrides the location of the user Nix configuration files
  to load from (defaults to the XDG spec locations). The variable is treated
  as a list separated by the <literal>:</literal> token.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>TMPDIR</envar></term>

  <listitem><para>Use the specified directory to store temporary
  files.  In particular, this includes temporary build directories;
  these can take up substantial amounts of disk space.  The default is
  <filename>/tmp</filename>.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook" xml:id="envar-remote"><term><envar>NIX_REMOTE</envar></term>

  <listitem><para>This variable should be set to
  <literal>daemon</literal> if you want to use the Nix daemon to
  execute Nix operations. This is necessary in <link linkend="ssec-multi-user">multi-user Nix installations</link>.
  If the Nix daemon's Unix socket is at some non-standard path,
  this variable should be set to <literal>unix://path/to/socket</literal>.
  Otherwise, it should be left unset.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_SHOW_STATS</envar></term>

  <listitem><para>If set to <literal>1</literal>, Nix will print some
  evaluation statistics, such as the number of values
  allocated.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>NIX_COUNT_CALLS</envar></term>

  <listitem><para>If set to <literal>1</literal>, Nix will print how
  often functions were called during Nix expression evaluation.  This
  is useful for profiling your Nix expressions.</para></listitem>

</varlistentry><varlistentry xmlns="http://docbook.org/ns/docbook"><term><envar>GC_INITIAL_HEAP_SIZE</envar></term>

  <listitem><para>If Nix has been configured to use the Boehm garbage
  collector, this variable sets the initial size of the heap in bytes.
  It defaults to 384 MiB.  Setting it to a low value reduces memory
  consumption, but will increase runtime due to the overhead of
  garbage collection.</para></listitem>

</varlistentry>
</variablelist>

</refsection>


</refentry>
<refentry xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-nix-prefetch-url">

<refmeta>
  <refentrytitle>nix-prefetch-url</refentrytitle>
  <manvolnum>1</manvolnum>
  <refmiscinfo class="source">Nix</refmiscinfo>
  <refmiscinfo class="version">3.0</refmiscinfo>
</refmeta>

<refnamediv>
  <refname>nix-prefetch-url</refname>
  <refpurpose>copy a file from a URL into the store and print its hash</refpurpose>
</refnamediv>

<refsynopsisdiv>
  <cmdsynopsis>
    <command>nix-prefetch-url</command>
    <arg><option>--version</option></arg>
    <arg><option>--type</option> <replaceable>hashAlgo</replaceable></arg>
    <arg><option>--print-path</option></arg>
    <arg><option>--unpack</option></arg>
    <arg><option>--name</option> <replaceable>name</replaceable></arg>
    <arg choice="plain"><replaceable>url</replaceable></arg>
    <arg><replaceable>hash</replaceable></arg>
  </cmdsynopsis>
</refsynopsisdiv>

<refsection><title>Description</title>

<para>The command <command>nix-prefetch-url</command> downloads the
file referenced by the URL <replaceable>url</replaceable>, prints its
cryptographic hash, and copies it into the Nix store.  The file name
in the store is
<filename><replaceable>hash</replaceable>-<replaceable>baseName</replaceable></filename>,
where <replaceable>baseName</replaceable> is everything following the
final slash in <replaceable>url</replaceable>.</para>

<para>This command is just a convenience for Nix expression writers.
Often a Nix expression fetches some source distribution from the
network using the <literal>fetchurl</literal> expression contained in
Nixpkgs.  However, <literal>fetchurl</literal> requires a
cryptographic hash.  If you don't know the hash, you would have to
download the file first, and then <literal>fetchurl</literal> would
download it again when you build your Nix expression.  Since
<literal>fetchurl</literal> uses the same name for the downloaded file
as <command>nix-prefetch-url</command>, the redundant download can be
avoided.</para>

<para>If <replaceable>hash</replaceable> is specified, then a download
is not performed if the Nix store already contains a file with the
same hash and base name.  Otherwise, the file is downloaded, and an
error is signaled if the actual hash of the file does not match the
specified hash.</para>

<para>This command prints the hash on standard output.  Additionally,
if the option <option>--print-path</option> is used, the path of the
downloaded file in the Nix store is also printed.</para>

</refsection>


<refsection><title>Options</title>

<variablelist>

  <varlistentry><term><option>--type</option> <replaceable>hashAlgo</replaceable></term>

    <listitem><para>Use the specified cryptographic hash algorithm,
    which can be one of <literal>md5</literal>,
    <literal>sha1</literal>, and
    <literal>sha256</literal>.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--print-path</option></term>

    <listitem><para>Print the store path of the downloaded file on
    standard output.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--unpack</option></term>

    <listitem><para>Unpack the archive (which must be a tarball or zip
    file) and add the result to the Nix store. The resulting hash can
    be used with functions such as Nixpkgs&#x2019;s
    <varname>fetchzip</varname> or
    <varname>fetchFromGitHub</varname>.</para></listitem>

  </varlistentry>

  <varlistentry><term><option>--name</option> <replaceable>name</replaceable></term>

    <listitem><para>Override the name of the file in the Nix store. By
    default, this is
    <literal><replaceable>hash</replaceable>-<replaceable>basename</replaceable></literal>,
    where <replaceable>basename</replaceable> is the last component of
    <replaceable>url</replaceable>. Overriding the name is necessary
    when <replaceable>basename</replaceable> contains characters that
    are not allowed in Nix store paths.</para></listitem>

  </varlistentry>

</variablelist>

</refsection>


<refsection><title>Examples</title>

<screen>
$ nix-prefetch-url ftp://ftp.gnu.org/pub/gnu/hello/hello-2.10.tar.gz
0ssi1wpaf7plaswqqjwigppsg5fyh99vdlb9kzl7c9lng89ndq1i

$ nix-prefetch-url --print-path mirror://gnu/hello/hello-2.10.tar.gz
0ssi1wpaf7plaswqqjwigppsg5fyh99vdlb9kzl7c9lng89ndq1i
/nix/store/3x7dwzq014bblazs7kq20p9hyzz0qh8g-hello-2.10.tar.gz

$ nix-prefetch-url --unpack --print-path https://github.com/NixOS/patchelf/archive/0.8.tar.gz
079agjlv0hrv7fxnx9ngipx14gyncbkllxrp9cccnh3a50fxcmy7
/nix/store/19zrmhm3m40xxaw81c8cqm6aljgrnwj2-0.8.tar.gz
</screen>

</refsection>


</refentry>

</chapter>
<chapter xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-files">

<title>Files</title>

<para>This section lists configuration files that you can use when you
work with Nix.</para>

<refentry xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" xml:id="sec-conf-file" version="5">

<refmeta>
  <refentrytitle>nix.conf</refentrytitle>
  <manvolnum>5</manvolnum>
  <refmiscinfo class="source">Nix</refmiscinfo>
  <refmiscinfo class="version">3.0</refmiscinfo>
</refmeta>

<refnamediv>
  <refname>nix.conf</refname>
  <refpurpose>Nix configuration file</refpurpose>
</refnamediv>

<refsection><title>Description</title>

<para>By default Nix reads settings from the following places:</para>

<para>The system-wide configuration file
<filename><replaceable>sysconfdir</replaceable>/nix/nix.conf</filename>
(i.e. <filename>/etc/nix/nix.conf</filename> on most systems), or
<filename>$NIX_CONF_DIR/nix.conf</filename> if
<envar>NIX_CONF_DIR</envar> is set. Values loaded in this file are not forwarded to the Nix daemon. The
client assumes that the daemon has already loaded them.
</para>

<para>User-specific configuration files:</para>

<para>
  If <envar>NIX_USER_CONF_FILES</envar> is set, then each path separated by
  <literal>:</literal> will be loaded in reverse order.
</para>

<para>
  Otherwise it will look for <filename>nix/nix.conf</filename> files in
  <envar>XDG_CONFIG_DIRS</envar> and <envar>XDG_CONFIG_HOME</envar>.

  The default location is <filename>$HOME/.config/nix.conf</filename> if
  those environment variables are unset.
</para>

<para>The configuration files consist of
<literal><replaceable>name</replaceable> =
<replaceable>value</replaceable></literal> pairs, one per line. Other
files can be included with a line like <literal>include
<replaceable>path</replaceable></literal>, where
<replaceable>path</replaceable> is interpreted relative to the current
conf file and a missing file is an error unless
<literal>!include</literal> is used instead.
Comments start with a <literal>#</literal> character.  Here is an
example configuration file:</para>

<programlisting>
keep-outputs = true       # Nice for developers
keep-derivations = true   # Idem
</programlisting>

<para>You can override settings on the command line using the
<option>--option</option> flag, e.g. <literal>--option keep-outputs
false</literal>.</para>

<para>The following settings are currently available:

<variablelist>


  <varlistentry xml:id="conf-allowed-uris"><term><literal>allowed-uris</literal></term>

    <listitem>

      <para>A list of URI prefixes to which access is allowed in
      restricted evaluation mode. For example, when set to
      <literal>https://github.com/NixOS</literal>, builtin functions
      such as <function>fetchGit</function> are allowed to access
      <literal>https://github.com/NixOS/patchelf.git</literal>.</para>

    </listitem>

  </varlistentry>


  <varlistentry xml:id="conf-allow-import-from-derivation"><term><literal>allow-import-from-derivation</literal></term>

    <listitem><para>By default, Nix allows you to <function>import</function> from a derivation,
    allowing building at evaluation time. With this option set to false, Nix will throw an error
    when evaluating an expression that uses this feature, allowing users to ensure their evaluation
    will not require any builds to take place.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-allow-new-privileges"><term><literal>allow-new-privileges</literal></term>

    <listitem><para>(Linux-specific.) By default, builders on Linux
    cannot acquire new privileges by calling setuid/setgid programs or
    programs that have file capabilities. For example, programs such
    as <command>sudo</command> or <command>ping</command> will
    fail. (Note that in sandbox builds, no such programs are available
    unless you bind-mount them into the sandbox via the
    <option>sandbox-paths</option> option.) You can allow the
    use of such programs by enabling this option. This is impure and
    usually undesirable, but may be useful in certain scenarios
    (e.g. to spin up containers or set up userspace network interfaces
    in tests).</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-allowed-users"><term><literal>allowed-users</literal></term>

    <listitem>

      <para>A list of names of users (separated by whitespace) that
      are allowed to connect to the Nix daemon. As with the
      <option>trusted-users</option> option, you can specify groups by
      prefixing them with <literal>@</literal>. Also, you can allow
      all users by specifying <literal>*</literal>. The default is
      <literal>*</literal>.</para>

      <para>Note that trusted users are always allowed to connect.</para>

    </listitem>

  </varlistentry>


  <varlistentry xml:id="conf-auto-optimise-store"><term><literal>auto-optimise-store</literal></term>

    <listitem><para>If set to <literal>true</literal>, Nix
    automatically detects files in the store that have identical
    contents, and replaces them with hard links to a single copy.
    This saves disk space.  If set to <literal>false</literal> (the
    default), you can still run <command>nix-store
    --optimise</command> to get rid of duplicate
    files.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="conf-builders">
    <term><literal>builders</literal></term>
    <listitem>
      <para>A list of machines on which to perform builds. <phrase condition="manual">See <xref linkend="chap-distributed-builds"/> for details.</phrase></para>
    </listitem>
  </varlistentry>


  <varlistentry xml:id="conf-builders-use-substitutes"><term><literal>builders-use-substitutes</literal></term>

    <listitem><para>If set to <literal>true</literal>, Nix will instruct
    remote build machines to use their own binary substitutes if available. In
    practical terms, this means that remote hosts will fetch as many build
    dependencies as possible from their own substitutes (e.g, from
    <literal>cache.nixos.org</literal>), instead of waiting for this host to
    upload them all. This can drastically reduce build times if the network
    connection between this computer and the remote build host is slow. Defaults
    to <literal>false</literal>.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="conf-build-users-group"><term><literal>build-users-group</literal></term>

    <listitem><para>This options specifies the Unix group containing
    the Nix build user accounts.  In multi-user Nix installations,
    builds should not be performed by the Nix account since that would
    allow users to arbitrarily modify the Nix store and database by
    supplying specially crafted builders; and they cannot be performed
    by the calling user since that would allow him/her to influence
    the build result.</para>

    <para>Therefore, if this option is non-empty and specifies a valid
    group, builds will be performed under the user accounts that are a
    member of the group specified here (as listed in
    <filename>/etc/group</filename>).  Those user accounts should not
    be used for any other purpose!</para>

    <para>Nix will never run two builds under the same user account at
    the same time.  This is to prevent an obvious security hole: a
    malicious user writing a Nix expression that modifies the build
    result of a legitimate Nix expression being built by another user.
    Therefore it is good to have as many Nix build user accounts as
    you can spare.  (Remember: uids are cheap.)</para>

    <para>The build users should have permission to create files in
    the Nix store, but not delete them.  Therefore,
    <filename>/nix/store</filename> should be owned by the Nix
    account, its group should be the group specified here, and its
    mode should be <literal>1775</literal>.</para>

    <para>If the build users group is empty, builds will be performed
    under the uid of the Nix process (that is, the uid of the caller
    if <envar>NIX_REMOTE</envar> is empty, the uid under which the Nix
    daemon runs if <envar>NIX_REMOTE</envar> is
    <literal>daemon</literal>).  Obviously, this should not be used in
    multi-user settings with untrusted users.</para>

    </listitem>

  </varlistentry>


  <varlistentry xml:id="conf-compress-build-log"><term><literal>compress-build-log</literal></term>

    <listitem><para>If set to <literal>true</literal> (the default),
    build logs written to <filename>/nix/var/log/nix/drvs</filename>
    will be compressed on the fly using bzip2.  Otherwise, they will
    not be compressed.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="conf-connect-timeout"><term><literal>connect-timeout</literal></term>

    <listitem>

      <para>The timeout (in seconds) for establishing connections in
      the binary cache substituter.  It corresponds to
      <command>curl</command>&#x2019;s <option>--connect-timeout</option>
      option.</para>

    </listitem>

  </varlistentry>


  <varlistentry xml:id="conf-cores"><term><literal>cores</literal></term>

    <listitem><para>Sets the value of the
    <envar>NIX_BUILD_CORES</envar> environment variable in the
    invocation of builders.  Builders can use this variable at their
    discretion to control the maximum amount of parallelism.  For
    instance, in Nixpkgs, if the derivation attribute
    <varname>enableParallelBuilding</varname> is set to
    <literal>true</literal>, the builder passes the
    <option>-j<replaceable>N</replaceable></option> flag to GNU Make.
    It can be overridden using the <option linkend="opt-cores">--cores</option> command line switch and
    defaults to <literal>1</literal>.  The value <literal>0</literal>
    means that the builder should use all available CPU cores in the
    system.</para>

    <para>See also <xref linkend="chap-tuning-cores-and-jobs"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-diff-hook"><term><literal>diff-hook</literal></term>
  <listitem>
    <para>
      Absolute path to an executable capable of diffing build results.
      The hook executes if <xref linkend="conf-run-diff-hook"/> is
      true, and the output of a build is known to not be the same.
      This program is not executed to determine if two results are the
      same.
    </para>

    <para>
      The diff hook is executed by the same user and group who ran the
      build. However, the diff hook does not have write access to the
      store path just built.
    </para>

    <para>The diff hook program receives three parameters:</para>

    <orderedlist>
      <listitem>
        <para>
          A path to the previous build's results
        </para>
      </listitem>

      <listitem>
        <para>
          A path to the current build's results
        </para>
      </listitem>

      <listitem>
        <para>
          The path to the build's derivation
        </para>
      </listitem>

      <listitem>
        <para>
          The path to the build's scratch directory. This directory
          will exist only if the build was run with
          <option>--keep-failed</option>.
        </para>
      </listitem>
    </orderedlist>

    <para>
      The stderr and stdout output from the diff hook will not be
      displayed to the user. Instead, it will print to the nix-daemon's
      log.
    </para>

    <para>When using the Nix daemon, <literal>diff-hook</literal> must
    be set in the <filename>nix.conf</filename> configuration file, and
    cannot be passed at the command line.
    </para>
  </listitem>
  </varlistentry>

  <varlistentry xml:id="conf-enforce-determinism">
    <term><literal>enforce-determinism</literal></term>

    <listitem><para>See <xref linkend="conf-repeat"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-extra-sandbox-paths">
    <term><literal>extra-sandbox-paths</literal></term>

    <listitem><para>A list of additional paths appended to
    <option>sandbox-paths</option>. Useful if you want to extend
    its default value.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-extra-platforms"><term><literal>extra-platforms</literal></term>

    <listitem><para>Platforms other than the native one which
    this machine is capable of building for. This can be useful for
    supporting additional architectures on compatible machines:
    i686-linux can be built on x86_64-linux machines (and the default
    for this setting reflects this); armv7 is backwards-compatible with
    armv6 and armv5tel; some aarch64 machines can also natively run
    32-bit ARM code; and qemu-user may be used to support non-native
    platforms (though this may be slow and buggy). Most values for this
    are not enabled by default because build systems will often
    misdetect the target platform and generate incompatible code, so you
    may wish to cross-check the results of using this option against
    proper natively-built versions of your
    derivations.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-extra-substituters"><term><literal>extra-substituters</literal></term>

    <listitem><para>Additional binary caches appended to those
    specified in <option>substituters</option>.  When used by
    unprivileged users, untrusted substituters (i.e. those not listed
    in <option>trusted-substituters</option>) are silently
    ignored.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="conf-fallback"><term><literal>fallback</literal></term>

    <listitem><para>If set to <literal>true</literal>, Nix will fall
    back to building from source if a binary substitute fails.  This
    is equivalent to the <option>--fallback</option> flag.  The
    default is <literal>false</literal>.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="conf-fsync-metadata"><term><literal>fsync-metadata</literal></term>

    <listitem><para>If set to <literal>true</literal>, changes to the
    Nix store metadata (in <filename>/nix/var/nix/db</filename>) are
    synchronously flushed to disk.  This improves robustness in case
    of system crashes, but reduces performance.  The default is
    <literal>true</literal>.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="conf-hashed-mirrors"><term><literal>hashed-mirrors</literal></term>

    <listitem><para>A list of web servers used by
    <function>builtins.fetchurl</function> to obtain files by hash.
    Given a hash type <replaceable>ht</replaceable> and a base-16 hash
    <replaceable>h</replaceable>, Nix will try to download the file
    from
    <literal>hashed-mirror/<replaceable>ht</replaceable>/<replaceable>h</replaceable></literal>.
    This allows files to be downloaded even if they have disappeared
    from their original URI. For example, given the hashed mirror
    <literal>http://tarballs.example.com/</literal>, when building the
    derivation

<programlisting>
builtins.fetchurl {
  url = "https://example.org/foo-1.2.3.tar.xz";
  sha256 = "2c26b46b68ffc68ff99b453c1d30413413422d706483bfa0f98a5e886266e7ae";
}
</programlisting>

    Nix will attempt to download this file from
    <literal>http://tarballs.example.com/sha256/2c26b46b68ffc68ff99b453c1d30413413422d706483bfa0f98a5e886266e7ae</literal>
    first. If it is not available there, if will try the original URI.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-http-connections"><term><literal>http-connections</literal></term>

    <listitem><para>The maximum number of parallel TCP connections
    used to fetch files from binary caches and by other downloads. It
    defaults to 25. 0 means no limit.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-keep-build-log"><term><literal>keep-build-log</literal></term>

    <listitem><para>If set to <literal>true</literal> (the default),
    Nix will write the build log of a derivation (i.e. the standard
    output and error of its builder) to the directory
    <filename>/nix/var/log/nix/drvs</filename>.  The build log can be
    retrieved using the command <command>nix-store -l
    <replaceable>path</replaceable></command>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-keep-derivations"><term><literal>keep-derivations</literal></term>

    <listitem><para>If <literal>true</literal> (default), the garbage
    collector will keep the derivations from which non-garbage store
    paths were built.  If <literal>false</literal>, they will be
    deleted unless explicitly registered as a root (or reachable from
    other roots).</para>

    <para>Keeping derivation around is useful for querying and
    traceability (e.g., it allows you to ask with what dependencies or
    options a store path was built), so by default this option is on.
    Turn it off to save a bit of disk space (or a lot if
    <literal>keep-outputs</literal> is also turned on).</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-keep-env-derivations"><term><literal>keep-env-derivations</literal></term>

    <listitem><para>If <literal>false</literal> (default), derivations
    are not stored in Nix user environments.  That is, the derivations of
    any build-time-only dependencies may be garbage-collected.</para>

    <para>If <literal>true</literal>, when you add a Nix derivation to
    a user environment, the path of the derivation is stored in the
    user environment.  Thus, the derivation will not be
    garbage-collected until the user environment generation is deleted
    (<command>nix-env --delete-generations</command>).  To prevent
    build-time-only dependencies from being collected, you should also
    turn on <literal>keep-outputs</literal>.</para>

    <para>The difference between this option and
    <literal>keep-derivations</literal> is that this one is
    &#x201C;sticky&#x201D;: it applies to any user environment created while this
    option was enabled, while <literal>keep-derivations</literal>
    only applies at the moment the garbage collector is
    run.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="conf-keep-outputs"><term><literal>keep-outputs</literal></term>

    <listitem><para>If <literal>true</literal>, the garbage collector
    will keep the outputs of non-garbage derivations.  If
    <literal>false</literal> (default), outputs will be deleted unless
    they are GC roots themselves (or reachable from other roots).</para>

    <para>In general, outputs must be registered as roots separately.
    However, even if the output of a derivation is registered as a
    root, the collector will still delete store paths that are used
    only at build time (e.g., the C compiler, or source tarballs
    downloaded from the network).  To prevent it from doing so, set
    this option to <literal>true</literal>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-max-build-log-size"><term><literal>max-build-log-size</literal></term>

    <listitem>

      <para>This option defines the maximum number of bytes that a
      builder can write to its stdout/stderr.  If the builder exceeds
      this limit, it&#x2019;s killed.  A value of <literal>0</literal> (the
      default) means that there is no limit.</para>

    </listitem>

  </varlistentry>

  <varlistentry xml:id="conf-max-free"><term><literal>max-free</literal></term>

    <listitem><para>When a garbage collection is triggered by the
    <literal>min-free</literal> option, it stops as soon as
    <literal>max-free</literal> bytes are available. The default is
    infinity (i.e. delete all garbage).</para></listitem>

  </varlistentry>

  <varlistentry xml:id="conf-max-jobs"><term><literal>max-jobs</literal></term>

    <listitem><para>This option defines the maximum number of jobs
    that Nix will try to build in parallel.  The default is
    <literal>1</literal>. The special value <literal>auto</literal>
    causes Nix to use the number of CPUs in your system.  <literal>0</literal>
    is useful when using remote builders to prevent any local builds (except for
    <literal>preferLocalBuild</literal> derivation attribute which executes locally
    regardless).  It can be
    overridden using the <option linkend="opt-max-jobs">--max-jobs</option> (<option>-j</option>)
    command line switch.</para>

    <para>See also <xref linkend="chap-tuning-cores-and-jobs"/>.</para>
    </listitem>
  </varlistentry>

  <varlistentry xml:id="conf-max-silent-time"><term><literal>max-silent-time</literal></term>

    <listitem>

      <para>This option defines the maximum number of seconds that a
      builder can go without producing any data on standard output or
      standard error.  This is useful (for instance in an automated
      build system) to catch builds that are stuck in an infinite
      loop, or to catch remote builds that are hanging due to network
      problems.  It can be overridden using the <option linkend="opt-max-silent-time">--max-silent-time</option> command
      line switch.</para>

      <para>The value <literal>0</literal> means that there is no
      timeout.  This is also the default.</para>

    </listitem>

  </varlistentry>

  <varlistentry xml:id="conf-min-free"><term><literal>min-free</literal></term>

    <listitem>
      <para>When free disk space in <filename>/nix/store</filename>
      drops below <literal>min-free</literal> during a build, Nix
      performs a garbage-collection until <literal>max-free</literal>
      bytes are available or there is no more garbage.  A value of
      <literal>0</literal> (the default) disables this feature.</para>
    </listitem>

  </varlistentry>

  <varlistentry xml:id="conf-narinfo-cache-negative-ttl"><term><literal>narinfo-cache-negative-ttl</literal></term>

    <listitem>

      <para>The TTL in seconds for negative lookups. If a store path is
      queried from a substituter but was not found, there will be a
      negative lookup cached in the local disk cache database for the
      specified duration.</para>

    </listitem>

  </varlistentry>

  <varlistentry xml:id="conf-narinfo-cache-positive-ttl"><term><literal>narinfo-cache-positive-ttl</literal></term>

    <listitem>

      <para>The TTL in seconds for positive lookups. If a store path is
      queried from a substituter, the result of the query will be cached
      in the local disk cache database including some of the NAR
      metadata. The default TTL is a month, setting a shorter TTL for
      positive lookups can be useful for binary caches that have
      frequent garbage collection, in which case having a more frequent
      cache invalidation would prevent trying to pull the path again and
      failing with a hash mismatch if the build isn't reproducible.
      </para>

    </listitem>

  </varlistentry>

  <varlistentry xml:id="conf-netrc-file"><term><literal>netrc-file</literal></term>

    <listitem><para>If set to an absolute path to a <filename>netrc</filename>
    file, Nix will use the HTTP authentication credentials in this file when
    trying to download from a remote host through HTTP or HTTPS. Defaults to
    <filename>$NIX_CONF_DIR/netrc</filename>.</para>

    <para>The <filename>netrc</filename> file consists of a list of
    accounts in the following format:

<screen>
machine <replaceable>my-machine</replaceable>
login <replaceable>my-username</replaceable>
password <replaceable>my-password</replaceable>
</screen>

    For the exact syntax, see <link xlink:href="https://ec.haxx.se/usingcurl-netrc.html">the
    <literal>curl</literal> documentation.</link></para>

    <note><para>This must be an absolute path, and <literal>~</literal>
    is not resolved. For example, <filename>~/.netrc</filename> won't
    resolve to your home directory's <filename>.netrc</filename>.</para></note>
    </listitem>

  </varlistentry>


  <varlistentry xml:id="conf-plugin-files">
    <term><literal>plugin-files</literal></term>
    <listitem>
      <para>
        A list of plugin files to be loaded by Nix. Each of these
        files will be dlopened by Nix, allowing them to affect
        execution through static initialization. In particular, these
        plugins may construct static instances of RegisterPrimOp to
        add new primops or constants to the expression language,
        RegisterStoreImplementation to add new store implementations,
        RegisterCommand to add new subcommands to the
        <literal>nix</literal> command, and RegisterSetting to add new
        nix config settings. See the constructors for those types for
        more details.
      </para>
      <para>
        Since these files are loaded into the same address space as
        Nix itself, they must be DSOs compatible with the instance of
        Nix running at the time (i.e. compiled against the same
        headers, not linked to any incompatible libraries). They
        should not be linked to any Nix libs directly, as those will
        be available already at load time.
      </para>
      <para>
        If an entry in the list is a directory, all files in the
        directory are loaded as plugins (non-recursively).
      </para>
    </listitem>

  </varlistentry>

  <varlistentry xml:id="conf-pre-build-hook"><term><literal>pre-build-hook</literal></term>

    <listitem>


      <para>If set, the path to a program that can set extra
      derivation-specific settings for this system. This is used for settings
      that can't be captured by the derivation model itself and are too variable
      between different versions of the same system to be hard-coded into nix.
      </para>

      <para>The hook is passed the derivation path and, if sandboxes are enabled,
      the sandbox directory. It can then modify the sandbox and send a series of
      commands to modify various settings to stdout. The currently recognized
      commands are:</para>

      <variablelist>
        <varlistentry xml:id="extra-sandbox-paths">
          <term><literal>extra-sandbox-paths</literal></term>

          <listitem>

            <para>Pass a list of files and directories to be included in the
            sandbox for this build. One entry per line, terminated by an empty
            line. Entries have the same format as
            <literal>sandbox-paths</literal>.</para>

          </listitem>

        </varlistentry>
      </variablelist>
    </listitem>

  </varlistentry>

  <varlistentry xml:id="conf-post-build-hook">
    <term><literal>post-build-hook</literal></term>
    <listitem>
      <para>Optional. The path to a program to execute after each build.</para>

      <para>This option is only settable in the global
      <filename>nix.conf</filename>, or on the command line by trusted
      users.</para>

      <para>When using the nix-daemon, the daemon executes the hook as
      <literal>root</literal>. If the nix-daemon is not involved, the
      hook runs as the user executing the nix-build.</para>

      <itemizedlist>
        <listitem><para>The hook executes after an evaluation-time build.</para></listitem>
        <listitem><para>The hook does not execute on substituted paths.</para></listitem>
        <listitem><para>The hook's output always goes to the user's terminal.</para></listitem>
        <listitem><para>If the hook fails, the build succeeds but no further builds execute.</para></listitem>
        <listitem><para>The hook executes synchronously, and blocks other builds from progressing while it runs.</para></listitem>
      </itemizedlist>

      <para>The program executes with no arguments. The program's environment
      contains the following environment variables:</para>

      <variablelist>
        <varlistentry>
          <term><envar>DRV_PATH</envar></term>
          <listitem>
            <para>The derivation for the built paths.</para>
            <para>Example:
            <literal>/nix/store/5nihn1a7pa8b25l9zafqaqibznlvvp3f-bash-4.4-p23.drv</literal>
            </para>
          </listitem>
        </varlistentry>

        <varlistentry>
          <term><envar>OUT_PATHS</envar></term>
          <listitem>
            <para>Output paths of the built derivation, separated by a space character.</para>
            <para>Example:
            <literal>/nix/store/zf5lbh336mnzf1nlswdn11g4n2m8zh3g-bash-4.4-p23-dev
            /nix/store/rjxwxwv1fpn9wa2x5ssk5phzwlcv4mna-bash-4.4-p23-doc
            /nix/store/6bqvbzjkcp9695dq0dpl5y43nvy37pq1-bash-4.4-p23-info
            /nix/store/r7fng3kk3vlpdlh2idnrbn37vh4imlj2-bash-4.4-p23-man
            /nix/store/xfghy8ixrhz3kyy6p724iv3cxji088dx-bash-4.4-p23</literal>.
            </para>
          </listitem>
        </varlistentry>
      </variablelist>

      <para>See <xref linkend="chap-post-build-hook"/> for an example
      implementation.</para>

    </listitem>
  </varlistentry>

  <varlistentry xml:id="conf-repeat"><term><literal>repeat</literal></term>

    <listitem><para>How many times to repeat builds to check whether
    they are deterministic. The default value is 0. If the value is
    non-zero, every build is repeated the specified number of
    times. If the contents of any of the runs differs from the
    previous ones and <xref linkend="conf-enforce-determinism"/> is
    true, the build is rejected and the resulting store paths are not
    registered as &#x201C;valid&#x201D; in Nix&#x2019;s database.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-require-sigs"><term><literal>require-sigs</literal></term>

    <listitem><para>If set to <literal>true</literal> (the default),
    any non-content-addressed path added or copied to the Nix store
    (e.g. when substituting from a binary cache) must have a valid
    signature, that is, be signed using one of the keys listed in
    <option>trusted-public-keys</option> or
    <option>secret-key-files</option>. Set to <literal>false</literal>
    to disable signature checking.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-restrict-eval"><term><literal>restrict-eval</literal></term>

    <listitem>

      <para>If set to <literal>true</literal>, the Nix evaluator will
      not allow access to any files outside of the Nix search path (as
      set via the <envar>NIX_PATH</envar> environment variable or the
      <option>-I</option> option), or to URIs outside of
      <option>allowed-uri</option>. The default is
      <literal>false</literal>.</para>

    </listitem>

  </varlistentry>

  <varlistentry xml:id="conf-run-diff-hook"><term><literal>run-diff-hook</literal></term>
  <listitem>
    <para>
      If true, enable the execution of <xref linkend="conf-diff-hook"/>.
    </para>

    <para>
      When using the Nix daemon, <literal>run-diff-hook</literal> must
      be set in the <filename>nix.conf</filename> configuration file,
      and cannot be passed at the command line.
    </para>
  </listitem>
  </varlistentry>

  <varlistentry xml:id="conf-sandbox"><term><literal>sandbox</literal></term>

    <listitem><para>If set to <literal>true</literal>, builds will be
    performed in a <emphasis>sandboxed environment</emphasis>, i.e.,
    they&#x2019;re isolated from the normal file system hierarchy and will
    only see their dependencies in the Nix store, the temporary build
    directory, private versions of <filename>/proc</filename>,
    <filename>/dev</filename>, <filename>/dev/shm</filename> and
    <filename>/dev/pts</filename> (on Linux), and the paths configured with the
    <link linkend="conf-sandbox-paths"><literal>sandbox-paths</literal>
    option</link>. This is useful to prevent undeclared dependencies
    on files in directories such as <filename>/usr/bin</filename>. In
    addition, on Linux, builds run in private PID, mount, network, IPC
    and UTS namespaces to isolate them from other processes in the
    system (except that fixed-output derivations do not run in private
    network namespace to ensure they can access the network).</para>

    <para>Currently, sandboxing only work on Linux and macOS. The use
    of a sandbox requires that Nix is run as root (so you should use
    the <link linkend="conf-build-users-group">&#x201C;build users&#x201D;
    feature</link> to perform the actual builds under different users
    than root).</para>

    <para>If this option is set to <literal>relaxed</literal>, then
    fixed-output derivations and derivations that have the
    <varname>__noChroot</varname> attribute set to
    <literal>true</literal> do not run in sandboxes.</para>

    <para>The default is <literal>true</literal> on Linux and
    <literal>false</literal> on all other platforms.</para>

    </listitem>

  </varlistentry>

  <varlistentry xml:id="conf-sandbox-dev-shm-size"><term><literal>sandbox-dev-shm-size</literal></term>

    <listitem><para>This option determines the maximum size of the
    <literal>tmpfs</literal> filesystem mounted on
    <filename>/dev/shm</filename> in Linux sandboxes. For the format,
    see the description of the <option>size</option> option of
    <literal>tmpfs</literal> in
    <citerefentry><refentrytitle>mount</refentrytitle><manvolnum>8</manvolnum></citerefentry>. The
    default is <literal>50%</literal>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-sandbox-paths">
    <term><literal>sandbox-paths</literal></term>

    <listitem><para>A list of paths bind-mounted into Nix sandbox
    environments. You can use the syntax
    <literal><replaceable>target</replaceable>=<replaceable>source</replaceable></literal>
    to mount a path in a different location in the sandbox; for
    instance, <literal>/bin=/nix-bin</literal> will mount the path
    <literal>/nix-bin</literal> as <literal>/bin</literal> inside the
    sandbox. If <replaceable>source</replaceable> is followed by
    <literal>?</literal>, then it is not an error if
    <replaceable>source</replaceable> does not exist; for example,
    <literal>/dev/nvidiactl?</literal> specifies that
    <filename>/dev/nvidiactl</filename> will only be mounted in the
    sandbox if it exists in the host filesystem.</para>

    <para>Depending on how Nix was built, the default value for this option
    may be empty or provide <filename>/bin/sh</filename> as a
    bind-mount of <command>bash</command>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-secret-key-files"><term><literal>secret-key-files</literal></term>

    <listitem><para>A whitespace-separated list of files containing
    secret (private) keys. These are used to sign locally-built
    paths. They can be generated using <command>nix-store
    --generate-binary-cache-key</command>. The corresponding public
    key can be distributed to other users, who can add it to
    <option>trusted-public-keys</option> in their
    <filename>nix.conf</filename>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-show-trace"><term><literal>show-trace</literal></term>

    <listitem><para>Causes Nix to print out a stack trace in case of Nix
    expression evaluation errors.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-substitute"><term><literal>substitute</literal></term>

    <listitem><para>If set to <literal>true</literal> (default), Nix
    will use binary substitutes if available.  This option can be
    disabled to force building from source.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="conf-stalled-download-timeout"><term><literal>stalled-download-timeout</literal></term>
    <listitem>
      <para>The timeout (in seconds) for receiving data from servers
      during download. Nix cancels idle downloads after this timeout's
      duration.</para>
    </listitem>
  </varlistentry>

  <varlistentry xml:id="conf-substituters"><term><literal>substituters</literal></term>

    <listitem><para>A list of URLs of substituters, separated by
    whitespace.  The default is
    <literal>https://cache.nixos.org</literal>.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="conf-system"><term><literal>system</literal></term>

    <listitem><para>This option specifies the canonical Nix system
    name of the current installation, such as
    <literal>i686-linux</literal> or
    <literal>x86_64-darwin</literal>.  Nix can only build derivations
    whose <literal>system</literal> attribute equals the value
    specified here.  In general, it never makes sense to modify this
    value from its default, since you can use it to &#x2018;lie&#x2019; about the
    platform you are building on (e.g., perform a Mac OS build on a
    Linux machine; the result would obviously be wrong).  It only
    makes sense if the Nix binaries can run on multiple platforms,
    e.g., &#x2018;universal binaries&#x2019; that run on <literal>x86_64-linux</literal> and
    <literal>i686-linux</literal>.</para>

    <para>It defaults to the canonical Nix system name detected by
    <filename>configure</filename> at build time.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-system-features"><term><literal>system-features</literal></term>

    <listitem><para>A set of system &#x201C;features&#x201D; supported by this
    machine, e.g. <literal>kvm</literal>. Derivations can express a
    dependency on such features through the derivation attribute
    <varname>requiredSystemFeatures</varname>. For example, the
    attribute

<programlisting>
requiredSystemFeatures = [ "kvm" ];
</programlisting>

    ensures that the derivation can only be built on a machine with
    the <literal>kvm</literal> feature.</para>

    <para>This setting by default includes <literal>kvm</literal> if
    <filename>/dev/kvm</filename> is accessible, and the
    pseudo-features <literal>nixos-test</literal>,
    <literal>benchmark</literal> and <literal>big-parallel</literal>
    that are used in Nixpkgs to route builds to specific
    machines.</para>

    </listitem>

  </varlistentry>

  <varlistentry xml:id="conf-tarball-ttl"><term><literal>tarball-ttl</literal></term>

    <listitem>
      <para>Default: <literal>3600</literal> seconds.</para>

      <para>The number of seconds a downloaded tarball is considered
      fresh. If the cached tarball is stale, Nix will check whether
      it is still up to date using the ETag header. Nix will download
      a new version if the ETag header is unsupported, or the
      cached ETag doesn't match.
      </para>

      <para>Setting the TTL to <literal>0</literal> forces Nix to always
      check if the tarball is up to date.</para>

      <para>Nix caches tarballs in
      <filename>$XDG_CACHE_HOME/nix/tarballs</filename>.</para>

      <para>Files fetched via <envar>NIX_PATH</envar>,
      <function>fetchGit</function>, <function>fetchMercurial</function>,
      <function>fetchTarball</function>, and <function>fetchurl</function>
      respect this TTL.
      </para>
    </listitem>
  </varlistentry>

  <varlistentry xml:id="conf-timeout"><term><literal>timeout</literal></term>

    <listitem>

      <para>This option defines the maximum number of seconds that a
      builder can run.  This is useful (for instance in an automated
      build system) to catch builds that are stuck in an infinite loop
      but keep writing to their standard output or standard error.  It
      can be overridden using the <option linkend="opt-timeout">--timeout</option> command line
      switch.</para>

      <para>The value <literal>0</literal> means that there is no
      timeout.  This is also the default.</para>

    </listitem>

  </varlistentry>

  <varlistentry xml:id="conf-trace-function-calls"><term><literal>trace-function-calls</literal></term>

    <listitem>

      <para>Default: <literal>false</literal>.</para>

      <para>If set to <literal>true</literal>, the Nix evaluator will
      trace every function call. Nix will print a log message at the
      "vomit" level for every function entrance and function exit.</para>

      <informalexample><screen>
function-trace entered undefined position at 1565795816999559622
function-trace exited undefined position at 1565795816999581277
function-trace entered /nix/store/.../example.nix:226:41 at 1565795253249935150
function-trace exited /nix/store/.../example.nix:226:41 at 1565795253249941684
</screen></informalexample>

      <para>The <literal>undefined position</literal> means the function
      call is a builtin.</para>

      <para>Use the <literal>contrib/stack-collapse.py</literal> script
      distributed with the Nix source code to convert the trace logs
      in to a format suitable for <command>flamegraph.pl</command>.</para>

    </listitem>

  </varlistentry>

  <varlistentry xml:id="conf-trusted-public-keys"><term><literal>trusted-public-keys</literal></term>

    <listitem><para>A whitespace-separated list of public keys. When
    paths are copied from another Nix store (such as a binary cache),
    they must be signed with one of these keys. For example:
    <literal>cache.nixos.org-1:6NCHdD59X431o0gWypbMrAURkbJ16ZPMQFGspcDShjY=
    hydra.nixos.org-1:CNHJZBh9K4tP3EKF6FkkgeVYsS3ohTl+oS0Qa8bezVs=</literal>.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="conf-trusted-substituters"><term><literal>trusted-substituters</literal></term>

    <listitem><para>A list of URLs of substituters, separated by
    whitespace.  These are not used by default, but can be enabled by
    users of the Nix daemon by specifying <literal>--option
    substituters <replaceable>urls</replaceable></literal> on the
    command line.  Unprivileged users are only allowed to pass a
    subset of the URLs listed in <literal>substituters</literal> and
    <literal>trusted-substituters</literal>.</para></listitem>

  </varlistentry>

  <varlistentry xml:id="conf-trusted-users"><term><literal>trusted-users</literal></term>

    <listitem>

      <para>A list of names of users (separated by whitespace) that
      have additional rights when connecting to the Nix daemon, such
      as the ability to specify additional binary caches, or to import
      unsigned NARs. You can also specify groups by prefixing them
      with <literal>@</literal>; for instance,
      <literal>@wheel</literal> means all users in the
      <literal>wheel</literal> group. The default is
      <literal>root</literal>.</para>

      <warning><para>Adding a user to <option>trusted-users</option>
      is essentially equivalent to giving that user root access to the
      system. For example, the user can set
      <option>sandbox-paths</option> and thereby obtain read access to
      directories that are otherwise inacessible to
      them.</para></warning>

    </listitem>

  </varlistentry>

</variablelist>
</para>

<refsection>
  <title>Deprecated Settings</title>

<para>

<variablelist>

  <varlistentry xml:id="conf-binary-caches">
    <term><literal>binary-caches</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>binary-caches</literal> is now an alias to
    <xref linkend="conf-substituters"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-binary-cache-public-keys">
    <term><literal>binary-cache-public-keys</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>binary-cache-public-keys</literal> is now an alias to
    <xref linkend="conf-trusted-public-keys"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-build-compress-log">
    <term><literal>build-compress-log</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>build-compress-log</literal> is now an alias to
    <xref linkend="conf-compress-build-log"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-build-cores">
    <term><literal>build-cores</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>build-cores</literal> is now an alias to
    <xref linkend="conf-cores"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-build-extra-chroot-dirs">
    <term><literal>build-extra-chroot-dirs</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>build-extra-chroot-dirs</literal> is now an alias to
    <xref linkend="conf-extra-sandbox-paths"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-build-extra-sandbox-paths">
    <term><literal>build-extra-sandbox-paths</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>build-extra-sandbox-paths</literal> is now an alias to
    <xref linkend="conf-extra-sandbox-paths"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-build-fallback">
    <term><literal>build-fallback</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>build-fallback</literal> is now an alias to
    <xref linkend="conf-fallback"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-build-max-jobs">
    <term><literal>build-max-jobs</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>build-max-jobs</literal> is now an alias to
    <xref linkend="conf-max-jobs"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-build-max-log-size">
    <term><literal>build-max-log-size</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>build-max-log-size</literal> is now an alias to
    <xref linkend="conf-max-build-log-size"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-build-max-silent-time">
    <term><literal>build-max-silent-time</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>build-max-silent-time</literal> is now an alias to
    <xref linkend="conf-max-silent-time"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-build-repeat">
    <term><literal>build-repeat</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>build-repeat</literal> is now an alias to
    <xref linkend="conf-repeat"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-build-timeout">
    <term><literal>build-timeout</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>build-timeout</literal> is now an alias to
    <xref linkend="conf-timeout"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-build-use-chroot">
    <term><literal>build-use-chroot</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>build-use-chroot</literal> is now an alias to
    <xref linkend="conf-sandbox"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-build-use-sandbox">
    <term><literal>build-use-sandbox</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>build-use-sandbox</literal> is now an alias to
    <xref linkend="conf-sandbox"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-build-use-substitutes">
    <term><literal>build-use-substitutes</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>build-use-substitutes</literal> is now an alias to
    <xref linkend="conf-substitute"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-gc-keep-derivations">
    <term><literal>gc-keep-derivations</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>gc-keep-derivations</literal> is now an alias to
    <xref linkend="conf-keep-derivations"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-gc-keep-outputs">
    <term><literal>gc-keep-outputs</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>gc-keep-outputs</literal> is now an alias to
    <xref linkend="conf-keep-outputs"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-env-keep-derivations">
    <term><literal>env-keep-derivations</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>env-keep-derivations</literal> is now an alias to
    <xref linkend="conf-keep-env-derivations"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-extra-binary-caches">
    <term><literal>extra-binary-caches</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>extra-binary-caches</literal> is now an alias to
    <xref linkend="conf-extra-substituters"/>.</para></listitem>
  </varlistentry>

  <varlistentry xml:id="conf-trusted-binary-caches">
    <term><literal>trusted-binary-caches</literal></term>

    <listitem><para><emphasis>Deprecated:</emphasis>
    <literal>trusted-binary-caches</literal> is now an alias to
    <xref linkend="conf-trusted-substituters"/>.</para></listitem>
  </varlistentry>
</variablelist>
</para>
</refsection>

</refsection>

</refentry>

</chapter>

</part>
  <appendix xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xml:id="part-glossary" xml:base="glossary/glossary.xml">

<title>Glossary</title>


<glosslist>


<glossentry xml:id="gloss-derivation"><glossterm>derivation</glossterm>

  <glossdef><para>A description of a build action.  The result of a
  derivation is a store object.  Derivations are typically specified
  in Nix expressions using the <link linkend="ssec-derivation"><function>derivation</function>
  primitive</link>.  These are translated into low-level
  <emphasis>store derivations</emphasis> (implicitly by
  <command>nix-env</command> and <command>nix-build</command>, or
  explicitly by <command>nix-instantiate</command>).</para></glossdef>

</glossentry>


<glossentry><glossterm>store</glossterm>

  <glossdef><para>The location in the file system where store objects
  live.  Typically <filename>/nix/store</filename>.</para></glossdef>

</glossentry>


<glossentry><glossterm>store path</glossterm>

  <glossdef><para>The location in the file system of a store object,
  i.e., an immediate child of the Nix store
  directory.</para></glossdef>

</glossentry>


<glossentry><glossterm>store object</glossterm>

  <glossdef><para>A file that is an immediate child of the Nix store
  directory.  These can be regular files, but also entire directory
  trees.  Store objects can be sources (objects copied from outside of
  the store), derivation outputs (objects produced by running a build
  action), or derivations (files describing a build
  action).</para></glossdef>

</glossentry>


<glossentry xml:id="gloss-substitute"><glossterm>substitute</glossterm>

  <glossdef><para>A substitute is a command invocation stored in the
  Nix database that describes how to build a store object, bypassing
  the normal build mechanism (i.e., derivations).  Typically, the
  substitute builds the store object by downloading a pre-built
  version of the store object from some server.</para></glossdef>

</glossentry>


<glossentry><glossterm>purity</glossterm>

  <glossdef><para>The assumption that equal Nix derivations when run
  always produce the same output.  This cannot be guaranteed in
  general (e.g., a builder can rely on external inputs such as the
  network or the system time) but the Nix model assumes
  it.</para></glossdef>

</glossentry>


<glossentry><glossterm>Nix expression</glossterm>

  <glossdef><para>A high-level description of software packages and
  compositions thereof.  Deploying software using Nix entails writing
  Nix expressions for your packages.  Nix expressions are translated
  to derivations that are stored in the Nix store.  These derivations
  can then be built.</para></glossdef>

</glossentry>


<glossentry xml:id="gloss-reference"><glossterm>reference</glossterm>

  <glossdef>
    <para>A store path <varname>P</varname> is said to have a
    reference to a store path <varname>Q</varname> if the store object
    at <varname>P</varname> contains the path <varname>Q</varname>
    somewhere. The <emphasis>references</emphasis> of a store path are
    the set of store paths to which it has a reference.
    </para>
    <para>A derivation can reference other derivations and sources
    (but not output paths), whereas an output path only references other
    output paths.
    </para>
  </glossdef>

</glossentry>

<glossentry xml:id="gloss-reachable"><glossterm>reachable</glossterm>

  <glossdef><para>A store path <varname>Q</varname> is reachable from
  another store path <varname>P</varname> if <varname>Q</varname> is in the
  <link linkend="gloss-closure">closure</link> of the
  <link linkend="gloss-reference">references</link> relation.
  </para></glossdef>
</glossentry>

<glossentry xml:id="gloss-closure"><glossterm>closure</glossterm>

  <glossdef><para>The closure of a store path is the set of store
  paths that are directly or indirectly &#x201C;reachable&#x201D; from that store
  path; that is, it&#x2019;s the closure of the path under the <link linkend="gloss-reference">references</link> relation. For a package, the
  closure of its derivation is equivalent to the build-time
  dependencies, while the closure of its output path is equivalent to its
  runtime dependencies. For correct deployment it is necessary to deploy whole
  closures, since otherwise at runtime files could be missing. The command
  <command>nix-store -qR</command> prints out closures of store paths.
  </para>
  <para>As an example, if the store object at path <varname>P</varname> contains
  a reference to path <varname>Q</varname>, then <varname>Q</varname> is
  in the closure of <varname>P</varname>. Further, if <varname>Q</varname>
  references <varname>R</varname> then <varname>R</varname> is also in
  the closure of <varname>P</varname>.
  </para></glossdef>

</glossentry>


<glossentry xml:id="gloss-output-path"><glossterm>output path</glossterm>

  <glossdef><para>A store path produced by a derivation.</para></glossdef>

</glossentry>


<glossentry xml:id="gloss-deriver"><glossterm>deriver</glossterm>

  <glossdef><para>The deriver of an <link linkend="gloss-output-path">output path</link> is the store
  derivation that built it.</para></glossdef>

</glossentry>


<glossentry xml:id="gloss-validity"><glossterm>validity</glossterm>

  <glossdef><para>A store path is considered
  <emphasis>valid</emphasis> if it exists in the file system, is
  listed in the Nix database as being valid, and if all paths in its
  closure are also valid.</para></glossdef>

</glossentry>


<glossentry xml:id="gloss-user-env"><glossterm>user environment</glossterm>

  <glossdef><para>An automatically generated store object that
  consists of a set of symlinks to &#x201C;active&#x201D; applications, i.e., other
  store paths.  These are generated automatically by <link linkend="sec-nix-env"><command>nix-env</command></link>.  See <xref linkend="sec-profiles"/>.</para>

  </glossdef>

</glossentry>


<glossentry xml:id="gloss-profile"><glossterm>profile</glossterm>

  <glossdef><para>A symlink to the current <link linkend="gloss-user-env">user environment</link> of a user, e.g.,
  <filename>/nix/var/nix/profiles/default</filename>.</para></glossdef>

</glossentry>


<glossentry xml:id="gloss-nar"><glossterm>NAR</glossterm>

  <glossdef><para>A <emphasis>N</emphasis>ix
  <emphasis>AR</emphasis>chive.  This is a serialisation of a path in
  the Nix store.  It can contain regular files, directories and
  symbolic links.  NARs are generated and unpacked using
  <command>nix-store --dump</command> and <command>nix-store
  --restore</command>.</para></glossdef>

</glossentry>



</glosslist>


</appendix>
  <appendix xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xml:id="chap-hacking">

<title>Hacking</title>

<para>This section provides some notes on how to hack on Nix. To get
the latest version of Nix from GitHub:
<screen>
$ git clone https://github.com/NixOS/nix.git
$ cd nix
</screen>
</para>

<para>To build Nix for the current operating system/architecture use

<screen>
$ nix-build
</screen>

or if you have a flakes-enabled nix:

<screen>
$ nix build
</screen>

This will build <literal>defaultPackage</literal> attribute defined in the <literal>flake.nix</literal> file.

To build for other platforms add one of the following suffixes to it: aarch64-linux,
i686-linux, x86_64-darwin, x86_64-linux.

i.e.

<screen>
nix-build -A defaultPackage.x86_64-linux
</screen>

</para>

<para>To build all dependencies and start a shell in which all
environment variables are set up so that those dependencies can be
found:
<screen>
$ nix-shell
</screen>
To build Nix itself in this shell:
<screen>
[nix-shell]$ ./bootstrap.sh
[nix-shell]$ ./configure $configureFlags
[nix-shell]$ make -j $NIX_BUILD_CORES
</screen>
To install it in <literal>$(pwd)/inst</literal> and test it:
<screen>
[nix-shell]$ make install
[nix-shell]$ make installcheck
[nix-shell]$ ./inst/bin/nix --version
nix (Nix) 2.4
</screen>

If you have a flakes-enabled nix you can replace:

<screen>
$ nix-shell
</screen>

by:

<screen>
$ nix develop
</screen>

</para>

</appendix>
  <appendix xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="sec-relnotes" xml:base="release-notes/release-notes.xml">

<title>Nix Release Notes</title>

<!--
<partintro>
<para>This section lists the release notes for each stable version of Nix.</para>
</partintro>
-->

<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-2.3">

<title>Release 2.3 (2019-09-04)</title>

<para>This is primarily a bug fix release. However, it makes some
incompatible changes:</para>

<itemizedlist>

  <listitem>
    <para>Nix now uses BSD file locks instead of POSIX file
    locks. Because of this, you should not use Nix 2.3 and previous
    releases at the same time on a Nix store.</para>
  </listitem>

</itemizedlist>

<para>It also has the following changes:</para>

<itemizedlist>

  <listitem>
    <para><function>builtins.fetchGit</function>'s <varname>ref</varname>
    argument now allows specifying an absolute remote ref.
    Nix will automatically prefix <varname>ref</varname> with
    <literal>refs/heads</literal> only if <varname>ref</varname> doesn't
    already begin with <literal>refs/</literal>.
    </para>
  </listitem>

  <listitem>
    <para>The installer now enables sandboxing by default on Linux when the
    system has the necessary kernel support.
    </para>
  </listitem>

  <listitem>
    <para>The <literal>max-jobs</literal> setting now defaults to 1.</para>
  </listitem>

  <listitem>
    <para>New builtin functions:
    <literal>builtins.isPath</literal>,
    <literal>builtins.hashFile</literal>.
    </para>
  </listitem>

  <listitem>
    <para>The <command>nix</command> command has a new
    <option>--print-build-logs</option> (<option>-L</option>) flag to
    print build log output to stderr, rather than showing the last log
    line in the progress bar. To distinguish between concurrent
    builds, log lines are prefixed by the name of the package.
    </para>
  </listitem>

  <listitem>
    <para>Builds are now executed in a pseudo-terminal, and the
    <envar>TERM</envar> environment variable is set to
    <literal>xterm-256color</literal>. This allows many programs
    (e.g. <command>gcc</command>, <command>clang</command>,
    <command>cmake</command>) to print colorized log output.</para>
  </listitem>

  <listitem>
    <para>Add <option>--no-net</option> convenience flag. This flag
    disables substituters; sets the <literal>tarball-ttl</literal>
    setting to infinity (ensuring that any previously downloaded files
    are considered current); and disables retrying downloads and sets
    the connection timeout to the minimum. This flag is enabled
    automatically if there are no configured non-loopback network
    interfaces.</para>
  </listitem>

  <listitem>
    <para>Add a <literal>post-build-hook</literal> setting to run a
    program after a build has succeeded.</para>
  </listitem>

  <listitem>
    <para>Add a <literal>trace-function-calls</literal> setting to log
    the duration of Nix function calls to stderr.</para>
  </listitem>

</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-2.2">

<title>Release 2.2 (2019-01-11)</title>

<para>This is primarily a bug fix release. It also has the following
changes:</para>

<itemizedlist>

  <listitem>
    <para>In derivations that use structured attributes (i.e. that
    specify set the <varname>__structuredAttrs</varname> attribute to
    <literal>true</literal> to cause all attributes to be passed to
    the builder in JSON format), you can now specify closure checks
    per output, e.g.:

<programlisting>
outputChecks."out" = {
  # The closure of 'out' must not be larger than 256 MiB.
  maxClosureSize = 256 * 1024 * 1024;

  # It must not refer to C compiler or to the 'dev' output.
  disallowedRequisites = [ stdenv.cc "dev" ];
};

outputChecks."dev" = {
  # The 'dev' output must not be larger than 128 KiB.
  maxSize = 128 * 1024;
};
</programlisting>

    </para>
  </listitem>


  <listitem>
    <para>The derivation attribute
    <varname>requiredSystemFeatures</varname> is now enforced for
    local builds, and not just to route builds to remote builders.
    The supported features of a machine can be specified through the
    configuration setting <varname>system-features</varname>.</para>

    <para>By default, <varname>system-features</varname> includes
    <literal>kvm</literal> if <filename>/dev/kvm</filename>
    exists. For compatibility, it also includes the pseudo-features
    <literal>nixos-test</literal>, <literal>benchmark</literal> and
    <literal>big-parallel</literal> which are used by Nixpkgs to route
    builds to particular Hydra build machines.</para>

  </listitem>

  <listitem>
    <para>Sandbox builds are now enabled by default on Linux.</para>
  </listitem>

  <listitem>
    <para>The new command <command>nix doctor</command> shows
    potential issues with your Nix installation.</para>
  </listitem>

  <listitem>
    <para>The <literal>fetchGit</literal> builtin function now uses a
    caching scheme that puts different remote repositories in distinct
    local repositories, rather than a single shared repository. This
    may require more disk space but is faster.</para>
  </listitem>

  <listitem>
    <para>The <literal>dirOf</literal> builtin function now works on
    relative paths.</para>
  </listitem>

  <listitem>
    <para>Nix now supports <link xlink:href="https://www.w3.org/TR/SRI/">SRI hashes</link>,
    allowing the hash algorithm and hash to be specified in a single
    string. For example, you can write:

<programlisting>
import &lt;nix/fetchurl.nix&gt; {
  url = https://nixos.org/releases/nix/nix-2.1.3/nix-2.1.3.tar.xz;
  hash = "sha256-XSLa0FjVyADWWhFfkZ2iKTjFDda6mMXjoYMXLRSYQKQ=";
};
</programlisting>

    instead of

<programlisting>
import &lt;nix/fetchurl.nix&gt; {
  url = https://nixos.org/releases/nix/nix-2.1.3/nix-2.1.3.tar.xz;
  sha256 = "5d22dad058d5c800d65a115f919da22938c50dd6ba98c5e3a183172d149840a4";
};
</programlisting>

    </para>

    <para>In fixed-output derivations, the
    <varname>outputHashAlgo</varname> attribute is no longer mandatory
    if <varname>outputHash</varname> specifies the hash.</para>

    <para><command>nix hash-file</command> and <command>nix
    hash-path</command> now print hashes in SRI format by
    default. They also use SHA-256 by default instead of SHA-512
    because that's what we use most of the time in Nixpkgs.</para>
  </listitem>

  <listitem>
    <para>Integers are now 64 bits on all platforms.</para>
  </listitem>

  <listitem>
    <para>The evaluator now prints profiling statistics (enabled via
    the <envar>NIX_SHOW_STATS</envar> and
    <envar>NIX_COUNT_CALLS</envar> environment variables) in JSON
    format.</para>
  </listitem>

  <listitem>
    <para>The option <option>--xml</option> in <command>nix-store
    --query</command> has been removed. Instead, there now is an
    option <option>--graphml</option> to output the dependency graph
    in GraphML format.</para>
  </listitem>

  <listitem>
    <para>All <filename>nix-*</filename> commands are now symlinks to
    <filename>nix</filename>. This saves a bit of disk space.</para>
  </listitem>

  <listitem>
    <para><command>nix repl</command> now uses
    <literal>libeditline</literal> or
    <literal>libreadline</literal>.</para>
  </listitem>

</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-2.1">

<title>Release 2.1 (2018-09-02)</title>

<para>This is primarily a bug fix release. It also reduces memory
consumption in certain situations. In addition, it has the following
new features:</para>

<itemizedlist>

  <listitem>
    <para>The Nix installer will no longer default to the Multi-User
    installation for macOS. You can still <link linkend="sect-multi-user-installation">instruct the installer to
    run in multi-user mode</link>.
    </para>
  </listitem>

  <listitem>
    <para>The Nix installer now supports performing a Multi-User
    installation for Linux computers which are running systemd. You
    can <link linkend="sect-multi-user-installation">select a Multi-User installation</link> by passing the
    <option>--daemon</option> flag to the installer: <command>sh &lt;(curl
    https://nixos.org/nix/install) --daemon</command>.
    </para>

    <para>The multi-user installer cannot handle systems with SELinux.
    If your system has SELinux enabled, you can <link linkend="sect-single-user-installation">force the installer to run
    in single-user mode</link>.</para>
  </listitem>

  <listitem>
    <para>New builtin functions:
    <literal>builtins.bitAnd</literal>,
    <literal>builtins.bitOr</literal>,
    <literal>builtins.bitXor</literal>,
    <literal>builtins.fromTOML</literal>,
    <literal>builtins.concatMap</literal>,
    <literal>builtins.mapAttrs</literal>.
    </para>
  </listitem>

  <listitem>
    <para>The S3 binary cache store now supports uploading NARs larger
    than 5 GiB.</para>
  </listitem>

  <listitem>
    <para>The S3 binary cache store now supports uploading to
    S3-compatible services with the <literal>endpoint</literal>
    option.</para>
  </listitem>

  <listitem>
    <para>The flag <option>--fallback</option> is no longer required
    to recover from disappeared NARs in binary caches.</para>
  </listitem>

  <listitem>
    <para><command>nix-daemon</command> now respects
    <option>--store</option>.</para>
  </listitem>

  <listitem>
    <para><command>nix run</command> now respects
    <varname>nix-support/propagated-user-env-packages</varname>.</para>
  </listitem>

</itemizedlist>

<para>This release has contributions from

Adrien Devresse,
Aleksandr Pashkov,
Alexandre Esteves,
Amine Chikhaoui,
Andrew Dunham,
Asad Saeeduddin,
aszlig,
Ben Challenor,
Ben Gamari,
Benjamin Hipple,
Bogdan Seniuc,
Corey O'Connor,
Daiderd Jordan,
Daniel Peebles,
Daniel Poelzleithner,
Danylo Hlynskyi,
Dmitry Kalinkin,
Domen Ko&#x17E;ar,
Doug Beardsley,
Eelco Dolstra,
Erik Arvstedt,
F&#xE9;lix Baylac-Jacqu&#xE9;,
Gleb Peregud,
Graham Christensen,
Guillaume Maudoux,
Ivan Kozik,
John Arnold,
Justin Humm,
Linus Heckemann,
Lorenzo Manacorda,
Matthew Justin Bauer,
Matthew O'Gorman,
Maximilian Bosch,
Michael Bishop,
Michael Fiano,
Michael Mercier,
Michael Raskin,
Michael Weiss,
Nicolas Dudebout,
Peter Simons,
Ryan Trinkle,
Samuel Dionne-Riel,
Sean Seefried,
Shea Levy,
Symphorien Gibol,
Tim Engler,
Tim Sears,
Tuomas Tynkkynen,
volth,
Will Dietz,
Yorick van Pelt and
zimbatm.
</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-2.0">

<title>Release 2.0 (2018-02-22)</title>

<para>The following incompatible changes have been made:</para>

<itemizedlist>

  <listitem>
    <para>The manifest-based substituter mechanism
    (<command>download-using-manifests</command>) has been <link xlink:href="https://github.com/NixOS/nix/commit/867967265b80946dfe1db72d40324b4f9af988ed">removed</link>. It
    has been superseded by the binary cache substituter mechanism
    since several years. As a result, the following programs have been
    removed:

    <itemizedlist>
      <listitem><para><command>nix-pull</command></para></listitem>
      <listitem><para><command>nix-generate-patches</command></para></listitem>
      <listitem><para><command>bsdiff</command></para></listitem>
      <listitem><para><command>bspatch</command></para></listitem>
    </itemizedlist>
    </para>
  </listitem>

  <listitem>
    <para>The &#x201C;copy from other stores&#x201D; substituter mechanism
    (<command>copy-from-other-stores</command> and the
    <envar>NIX_OTHER_STORES</envar> environment variable) has been
    removed. It was primarily used by the NixOS installer to copy
    available paths from the installation medium. The replacement is
    to use a chroot store as a substituter
    (e.g. <literal>--substituters /mnt</literal>), or to build into a
    chroot store (e.g. <literal>--store /mnt --substituters /</literal>).</para>
  </listitem>

  <listitem>
    <para>The command <command>nix-push</command> has been removed as
    part of the effort to eliminate Nix's dependency on Perl. You can
    use <command>nix copy</command> instead, e.g. <literal>nix copy
    --to file:///tmp/my-binary-cache <replaceable>paths&#x2026;</replaceable></literal></para>
  </listitem>

  <listitem>
    <para>The &#x201C;nested&#x201D; log output feature (<option>--log-type
    pretty</option>) has been removed. As a result,
    <command>nix-log2xml</command> was also removed.</para>
  </listitem>

  <listitem>
    <para>OpenSSL-based signing has been <link xlink:href="https://github.com/NixOS/nix/commit/f435f8247553656774dd1b2c88e9de5d59cab203">removed</link>. This
    feature was never well-supported. A better alternative is provided
    by the <option>secret-key-files</option> and
    <option>trusted-public-keys</option> options.</para>
  </listitem>

  <listitem>
    <para>Failed build caching has been <link xlink:href="https://github.com/NixOS/nix/commit/8cffec84859cec8b610a2a22ab0c4d462a9351ff">removed</link>. This
    feature was introduced to support the Hydra continuous build
    system, but Hydra no longer uses it.</para>
  </listitem>

  <listitem>
    <para><filename>nix-mode.el</filename> has been removed from
    Nix. It is now <link xlink:href="https://github.com/NixOS/nix-mode">a separate
    repository</link> and can be installed through the MELPA package
    repository.</para>
  </listitem>

</itemizedlist>

<para>This release has the following new features:</para>

<itemizedlist>

  <listitem>
    <para>It introduces a new command named <command>nix</command>,
    which is intended to eventually replace all
    <command>nix-*</command> commands with a more consistent and
    better designed user interface. It currently provides replacements
    for some (but not all) of the functionality provided by
    <command>nix-store</command>, <command>nix-build</command>,
    <command>nix-shell -p</command>, <command>nix-env -qa</command>,
    <command>nix-instantiate --eval</command>,
    <command>nix-push</command> and
    <command>nix-copy-closure</command>. It has the following major
    features:</para>

    <itemizedlist>

      <listitem>
        <para>Unlike the legacy commands, it has a consistent way to
        refer to packages and package-like arguments (like store
        paths). For example, the following commands all copy the GNU
        Hello package to a remote machine:

        <screen>nix copy --to ssh://machine nixpkgs.hello</screen>
        <screen>nix copy --to ssh://machine /nix/store/0i2jd68mp5g6h2sa5k9c85rb80sn8hi9-hello-2.10</screen>
        <screen>nix copy --to ssh://machine '(with import &lt;nixpkgs&gt; {}; hello)'</screen>

        By contrast, <command>nix-copy-closure</command> only accepted
        store paths as arguments.</para>
      </listitem>

      <listitem>
        <para>It is self-documenting: <option>--help</option> shows
        all available command-line arguments. If
        <option>--help</option> is given after a subcommand, it shows
        examples for that subcommand. <command>nix
        --help-config</command> shows all configuration
        options.</para>
      </listitem>

      <listitem>
        <para>It is much less verbose. By default, it displays a
        single-line progress indicator that shows how many packages
        are left to be built or downloaded, and (if there are running
        builds) the most recent line of builder output. If a build
        fails, it shows the last few lines of builder output. The full
        build log can be retrieved using <command>nix
        log</command>.</para>
      </listitem>

      <listitem>
        <para>It <link xlink:href="https://github.com/NixOS/nix/commit/b8283773bd64d7da6859ed520ee19867742a03ba">provides</link>
        all <filename>nix.conf</filename> configuration options as
        command line flags. For example, instead of <literal>--option
        http-connections 100</literal> you can write
        <literal>--http-connections 100</literal>. Boolean options can
        be written as
        <literal>--<replaceable>foo</replaceable></literal> or
        <literal>--no-<replaceable>foo</replaceable></literal>
        (e.g. <option>--no-auto-optimise-store</option>).</para>
      </listitem>

      <listitem>
        <para>Many subcommands have a <option>--json</option> flag to
        write results to stdout in JSON format.</para>
      </listitem>

    </itemizedlist>

    <warning><para>Please note that the <command>nix</command> command
    is a work in progress and the interface is subject to
    change.</para></warning>

    <para>It provides the following high-level (&#x201C;porcelain&#x201D;)
    subcommands:</para>

    <itemizedlist>

      <listitem>
        <para><command>nix build</command> is a replacement for
        <command>nix-build</command>.</para>
      </listitem>

      <listitem>
        <para><command>nix run</command> executes a command in an
        environment in which the specified packages are available. It
        is (roughly) a replacement for <command>nix-shell
        -p</command>. Unlike that command, it does not execute the
        command in a shell, and has a flag (<command>-c</command>)
        that specifies the unquoted command line to be
        executed.</para>

        <para>It is particularly useful in conjunction with chroot
        stores, allowing Linux users who do not have permission to
        install Nix in <command>/nix/store</command> to still use
        binary substitutes that assume
        <command>/nix/store</command>. For example,

        <screen>nix run --store ~/my-nix nixpkgs.hello -c hello --greeting 'Hi everybody!'</screen>

        downloads (or if not substitutes are available, builds) the
        GNU Hello package into
        <filename>~/my-nix/nix/store</filename>, then runs
        <command>hello</command> in a mount namespace where
        <filename>~/my-nix/nix/store</filename> is mounted onto
        <command>/nix/store</command>.</para>
      </listitem>

      <listitem>
        <para><command>nix search</command> replaces <command>nix-env
        -qa</command>. It searches the available packages for
        occurrences of a search string in the attribute name, package
        name or description. Unlike <command>nix-env -qa</command>, it
        has a cache to speed up subsequent searches.</para>
      </listitem>

      <listitem>
        <para><command>nix copy</command> copies paths between
        arbitrary Nix stores, generalising
        <command>nix-copy-closure</command> and
        <command>nix-push</command>.</para>
      </listitem>

      <listitem>
        <para><command>nix repl</command> replaces the external
        program <command>nix-repl</command>. It provides an
        interactive environment for evaluating and building Nix
        expressions. Note that it uses <literal>linenoise-ng</literal>
        instead of GNU Readline.</para>
      </listitem>

      <listitem>
        <para><command>nix upgrade-nix</command> upgrades Nix to the
        latest stable version. This requires that Nix is installed in
        a profile. (Thus it won&#x2019;t work on NixOS, or if it&#x2019;s installed
        outside of the Nix store.)</para>
      </listitem>

      <listitem>
        <para><command>nix verify</command> checks whether store paths
        are unmodified and/or &#x201C;trusted&#x201D; (see below). It replaces
        <command>nix-store --verify</command> and <command>nix-store
        --verify-path</command>.</para>
      </listitem>

      <listitem>
        <para><command>nix log</command> shows the build log of a
        package or path. If the build log is not available locally, it
        will try to obtain it from the configured substituters (such
        as <uri>cache.nixos.org</uri>, which now provides build
        logs).</para>
      </listitem>

      <listitem>
        <para><command>nix edit</command> opens the source code of a
        package in your editor.</para>
      </listitem>

      <listitem>
        <para><command>nix eval</command> replaces
        <command>nix-instantiate --eval</command>.</para>
      </listitem>

      <listitem>
        <para><command xlink:href="https://github.com/NixOS/nix/commit/d41c5eb13f4f3a37d80dbc6d3888644170c3b44a">nix
        why-depends</command> shows why one store path has another in
        its closure. This is primarily useful to finding the causes of
        closure bloat. For example,

        <screen>nix why-depends nixpkgs.vlc nixpkgs.libdrm.dev</screen>

        shows a chain of files and fragments of file contents that
        cause the VLC package to have the &#x201C;dev&#x201D; output of
        <literal>libdrm</literal> in its closure &#x2014; an undesirable
        situation.</para>
      </listitem>

      <listitem>
        <para><command>nix path-info</command> shows information about
        store paths, replacing <command>nix-store -q</command>. A
        useful feature is the option <option>--closure-size</option>
        (<option>-S</option>). For example, the following command show
        the closure sizes of every path in the current NixOS system
        closure, sorted by size:

        <screen>nix path-info -rS /run/current-system | sort -nk2</screen>

        </para>
      </listitem>

      <listitem>
        <para><command>nix optimise-store</command> replaces
        <command>nix-store --optimise</command>. The main difference
        is that it has a progress indicator.</para>
      </listitem>

    </itemizedlist>

    <para>A number of low-level (&#x201C;plumbing&#x201D;) commands are also
    available:</para>

    <itemizedlist>

      <listitem>
        <para><command>nix ls-store</command> and <command>nix
        ls-nar</command> list the contents of a store path or NAR
        file. The former is primarily useful in conjunction with
        remote stores, e.g.

        <screen>nix ls-store --store https://cache.nixos.org/ -lR /nix/store/0i2jd68mp5g6h2sa5k9c85rb80sn8hi9-hello-2.10</screen>

        lists the contents of path in a binary cache.</para>
      </listitem>

      <listitem>
        <para><command>nix cat-store</command> and <command>nix
        cat-nar</command> allow extracting a file from a store path or
        NAR file.</para>
      </listitem>

      <listitem>
        <para><command>nix dump-path</command> writes the contents of
        a store path to stdout in NAR format. This replaces
        <command>nix-store --dump</command>.</para>
      </listitem>

      <listitem>
        <para><command xlink:href="https://github.com/NixOS/nix/commit/e8d6ee7c1b90a2fe6d824f1a875acc56799ae6e2">nix
        show-derivation</command> displays a store derivation in JSON
        format. This is an alternative to
        <command>pp-aterm</command>.</para>
      </listitem>

      <listitem>
        <para><command xlink:href="https://github.com/NixOS/nix/commit/970366266b8df712f5f9cedb45af183ef5a8357f">nix
        add-to-store</command> replaces <command>nix-store
        --add</command>.</para>
      </listitem>

      <listitem>
        <para><command>nix sign-paths</command> signs store
        paths.</para>
      </listitem>

      <listitem>
        <para><command>nix copy-sigs</command> copies signatures from
        one store to another.</para>
      </listitem>

      <listitem>
        <para><command>nix show-config</command> shows all
        configuration options and their current values.</para>
      </listitem>

    </itemizedlist>

  </listitem>

  <listitem>
    <para>The store abstraction that Nix has had for a long time to
    support store access via the Nix daemon has been extended
    significantly. In particular, substituters (which used to be
    external programs such as
    <command>download-from-binary-cache</command>) are now subclasses
    of the abstract <classname>Store</classname> class. This allows
    many Nix commands to operate on such store types. For example,
    <command>nix path-info</command> shows information about paths in
    your local Nix store, while <command>nix path-info --store
    https://cache.nixos.org/</command> shows information about paths
    in the specified binary cache. Similarly,
    <command>nix-copy-closure</command>, <command>nix-push</command>
    and substitution are all instances of the general notion of
    copying paths between different kinds of Nix stores.</para>

    <para>Stores are specified using an URI-like syntax,
    e.g. <uri>https://cache.nixos.org/</uri> or
    <uri>ssh://machine</uri>. The following store types are supported:

    <itemizedlist>

      <listitem>

        <para><classname>LocalStore</classname> (stori URI
        <literal>local</literal> or an absolute path) and the misnamed
        <classname>RemoteStore</classname> (<literal>daemon</literal>)
        provide access to a local Nix store, the latter via the Nix
        daemon. You can use <literal>auto</literal> or the empty
        string to auto-select a local or daemon store depending on
        whether you have write permission to the Nix store. It is no
        longer necessary to set the <envar>NIX_REMOTE</envar>
        environment variable to use the Nix daemon.</para>

        <para>As noted above, <classname>LocalStore</classname> now
        supports chroot builds, allowing the &#x201C;physical&#x201D; location of
        the Nix store
        (e.g. <filename>/home/alice/nix/store</filename>) to differ
        from its &#x201C;logical&#x201D; location (typically
        <filename>/nix/store</filename>). This allows non-root users
        to use Nix while still getting the benefits from prebuilt
        binaries from <uri>cache.nixos.org</uri>.</para>

      </listitem>

      <listitem>

        <para><classname>BinaryCacheStore</classname> is the abstract
        superclass of all binary cache stores. It supports writing
        build logs and NAR content listings in JSON format.</para>

      </listitem>

      <listitem>

        <para><classname>HttpBinaryCacheStore</classname>
        (<literal>http://</literal>, <literal>https://</literal>)
        supports binary caches via HTTP or HTTPS. If the server
        supports <literal>PUT</literal> requests, it supports
        uploading store paths via commands such as <command>nix
        copy</command>.</para>

      </listitem>

      <listitem>

        <para><classname>LocalBinaryCacheStore</classname>
        (<literal>file://</literal>) supports binary caches in the
        local filesystem.</para>

      </listitem>

      <listitem>

        <para><classname>S3BinaryCacheStore</classname>
        (<literal>s3://</literal>) supports binary caches stored in
        Amazon S3, if enabled at compile time.</para>

      </listitem>

      <listitem>

        <para><classname>LegacySSHStore</classname> (<literal>ssh://</literal>)
        is used to implement remote builds and
        <command>nix-copy-closure</command>.</para>

      </listitem>

      <listitem>

        <para><classname>SSHStore</classname>
        (<literal>ssh-ng://</literal>) supports arbitrary Nix
        operations on a remote machine via the same protocol used by
        <command>nix-daemon</command>.</para>

      </listitem>

    </itemizedlist>

    </para>

  </listitem>

  <listitem>

    <para>Security has been improved in various ways:

    <itemizedlist>

      <listitem>
        <para>Nix now stores signatures for local store
        paths. When paths are copied between stores (e.g., copied from
        a binary cache to a local store), signatures are
        propagated.</para>

        <para>Locally-built paths are signed automatically using the
        secret keys specified by the <option>secret-key-files</option>
        store option. Secret/public key pairs can be generated using
        <command>nix-store
        --generate-binary-cache-key</command>.</para>

        <para>In addition, locally-built store paths are marked as
        &#x201C;ultimately trusted&#x201D;, but this bit is not propagated when
        paths are copied between stores.</para>
      </listitem>

      <listitem>
        <para>Content-addressable store paths no longer require
        signatures &#x2014; they can be imported into a store by unprivileged
        users even if they lack signatures.</para>
      </listitem>

      <listitem>
        <para>The command <command>nix verify</command> checks whether
        the specified paths are trusted, i.e., have a certain number
        of trusted signatures, are ultimately trusted, or are
        content-addressed.</para>
      </listitem>

      <listitem>
        <para>Substitutions from binary caches <link xlink:href="https://github.com/NixOS/nix/commit/ecbc3fedd3d5bdc5a0e1a0a51b29062f2874ac8b">now</link>
        require signatures by default. This was already the case on
        NixOS.</para>
      </listitem>

      <listitem>
        <para>In Linux sandbox builds, we <link xlink:href="https://github.com/NixOS/nix/commit/eba840c8a13b465ace90172ff76a0db2899ab11b">now</link>
        use <filename>/build</filename> instead of
        <filename>/tmp</filename> as the temporary build
        directory. This fixes potential security problems when a build
        accidentally stores its <envar>TMPDIR</envar> in some
        security-sensitive place, such as an RPATH.</para>
      </listitem>

    </itemizedlist>

    </para>

  </listitem>

  <listitem>
    <para><emphasis>Pure evaluation mode</emphasis>. With the
    <literal>--pure-eval</literal> flag, Nix enables a variant of the existing
    restricted evaluation mode that forbids access to anything that could cause
    different evaluations of the same command line arguments to produce a
    different result. This includes builtin functions such as
    <function>builtins.getEnv</function>, but more importantly,
    <emphasis>all</emphasis> filesystem or network access unless a content hash
    or commit hash is specified. For example, calls to
    <function>builtins.fetchGit</function> are only allowed if a
    <varname>rev</varname> attribute is specified.</para>

    <para>The goal of this feature is to enable true reproducibility
    and traceability of builds (including NixOS system configurations)
    at the evaluation level. For example, in the future,
    <command>nixos-rebuild</command> might build configurations from a
    Nix expression in a Git repository in pure mode. That expression
    might fetch other repositories such as Nixpkgs via
    <function>builtins.fetchGit</function>. The commit hash of the
    top-level repository then uniquely identifies a running system,
    and, in conjunction with that repository, allows it to be
    reproduced or modified.</para>

  </listitem>

  <listitem>
    <para>There are several new features to support binary
    reproducibility (i.e. to help ensure that multiple builds of the
    same derivation produce exactly the same output). When
    <option>enforce-determinism</option> is set to
    <literal>false</literal>, it&#x2019;s <link xlink:href="https://github.com/NixOS/nix/commit/8bdf83f936adae6f2c907a6d2541e80d4120f051">no
    longer</link> a fatal error if build rounds produce different
    output. Also, a hook named <option>diff-hook</option> is <link xlink:href="https://github.com/NixOS/nix/commit/9a313469a4bdea2d1e8df24d16289dc2a172a169">provided</link>
    to allow you to run tools such as <command>diffoscope</command>
    when build rounds produce different output.</para>
  </listitem>

  <listitem>
    <para>Configuring remote builds is a lot easier now. Provided you
    are not using the Nix daemon, you can now just specify a remote
    build machine on the command line, e.g. <literal>--option builders
    'ssh://my-mac x86_64-darwin'</literal>. The environment variable
    <envar>NIX_BUILD_HOOK</envar> has been removed and is no longer
    needed. The environment variable <envar>NIX_REMOTE_SYSTEMS</envar>
    is still supported for compatibility, but it is also possible to
    specify builders in <command>nix.conf</command> by setting the
    option <literal>builders =
    @<replaceable>path</replaceable></literal>.</para>
  </listitem>

  <listitem>
    <para>If a fixed-output derivation produces a result with an
    incorrect hash, the output path is moved to the location
    corresponding to the actual hash and registered as valid. Thus, a
    subsequent build of the fixed-output derivation with the correct
    hash is unnecessary.</para>
  </listitem>

  <listitem>
    <para><command>nix-shell</command> <link xlink:href="https://github.com/NixOS/nix/commit/ea59f39326c8e9dc42dfed4bcbf597fbce58797c">now</link>
    sets the <varname>IN_NIX_SHELL</varname> environment variable
    during evaluation and in the shell itself. This can be used to
    perform different actions depending on whether you&#x2019;re in a Nix
    shell or in a regular build. Nixpkgs provides
    <varname>lib.inNixShell</varname> to check this variable during
    evaluation.</para>
  </listitem>

  <listitem>
    <para><envar>NIX_PATH</envar> is now lazy, so URIs in the path are
    only downloaded if they are needed for evaluation.</para>
  </listitem>

  <listitem>
    <para>You can now use
    <uri>channel:<replaceable>channel-name</replaceable></uri> as a
    short-hand for
    <uri>https://nixos.org/channels/<replaceable>channel-name</replaceable>/nixexprs.tar.xz</uri>. For
    example, <literal>nix-build channel:nixos-15.09 -A hello</literal>
    will build the GNU Hello package from the
    <literal>nixos-15.09</literal> channel. In the future, this may
    use Git to fetch updates more efficiently.</para>
  </listitem>

  <listitem>
    <para>When <option>--no-build-output</option> is given, the last
    10 lines of the build log will be shown if a build
    fails.</para>
  </listitem>

  <listitem>
    <para>Networking has been improved:

    <itemizedlist>

      <listitem>
        <para>HTTP/2 is now supported. This makes binary cache lookups
        <link xlink:href="https://github.com/NixOS/nix/commit/90ad02bf626b885a5dd8967894e2eafc953bdf92">much
        more efficient</link>.</para>
      </listitem>

      <listitem>
        <para>We now retry downloads on many HTTP errors, making
        binary caches substituters more resilient to temporary
        failures.</para>
      </listitem>

      <listitem>
        <para>HTTP credentials can now be configured via the standard
        <filename>netrc</filename> mechanism.</para>
      </listitem>

      <listitem>
        <para>If S3 support is enabled at compile time,
        <uri>s3://</uri> URIs are <link xlink:href="https://github.com/NixOS/nix/commit/9ff9c3f2f80ba4108e9c945bbfda2c64735f987b">supported</link>
        in all places where Nix allows URIs.</para>
      </listitem>

      <listitem>
        <para>Brotli compression is now supported. In particular,
        <uri>cache.nixos.org</uri> build logs are now compressed using
        Brotli.</para>
      </listitem>

    </itemizedlist>

    </para>

  </listitem>

  <listitem>
    <para><command>nix-env</command> <link xlink:href="https://github.com/NixOS/nix/commit/b0cb11722626e906a73f10dd9a0c9eea29faf43a">now</link>
    ignores packages with bad derivation names (in particular those
    starting with a digit or containing a dot).</para>
  </listitem>

  <listitem>
    <para>Many configuration options have been renamed, either because
    they were unnecessarily verbose
    (e.g. <option>build-use-sandbox</option> is now just
    <option>sandbox</option>) or to reflect generalised behaviour
    (e.g. <option>binary-caches</option> is now
    <option>substituters</option> because it allows arbitrary store
    URIs). The old names are still supported for compatibility.</para>
  </listitem>

  <listitem>
    <para>The <option>max-jobs</option> option can <link xlink:href="https://github.com/NixOS/nix/commit/7251d048fa812d2551b7003bc9f13a8f5d4c95a5">now</link>
    be set to <literal>auto</literal> to use the number of CPUs in the
    system.</para>
  </listitem>

  <listitem>
    <para>Hashes can <link xlink:href="https://github.com/NixOS/nix/commit/c0015e87af70f539f24d2aa2bc224a9d8b84276b">now</link>
    be specified in base-64 format, in addition to base-16 and the
    non-standard base-32.</para>
  </listitem>

  <listitem>
    <para><command>nix-shell</command> now uses
    <varname>bashInteractive</varname> from Nixpkgs, rather than the
    <command>bash</command> command that happens to be in the caller&#x2019;s
    <envar>PATH</envar>. This is especially important on macOS where
    the <command>bash</command> provided by the system is seriously
    outdated and cannot execute <literal>stdenv</literal>&#x2019;s setup
    script.</para>
  </listitem>

  <listitem>
    <para>Nix can now automatically trigger a garbage collection if
    free disk space drops below a certain level during a build. This
    is configured using the <option>min-free</option> and
    <option>max-free</option> options.</para>
  </listitem>

  <listitem>
    <para><command>nix-store -q --roots</command> and
    <command>nix-store --gc --print-roots</command> now show temporary
    and in-memory roots.</para>
  </listitem>

  <listitem>
    <para>
      Nix can now be extended with plugins. See the documentation of
      the <option>plugin-files</option> option for more details.
    </para>
  </listitem>

</itemizedlist>

<para>The Nix language has the following new features:

<itemizedlist>

  <listitem>
    <para>It supports floating point numbers. They are based on the
    C++ <literal>float</literal> type and are supported by the
    existing numerical operators. Export and import to and from JSON
    and XML works, too.</para>
  </listitem>

  <listitem>
    <para>Derivation attributes can now reference the outputs of the
    derivation using the <function>placeholder</function> builtin
    function. For example, the attribute

<programlisting>
configureFlags = "--prefix=${placeholder "out"} --includedir=${placeholder "dev"}";
</programlisting>

    will cause the <envar>configureFlags</envar> environment variable
    to contain the actual store paths corresponding to the
    <literal>out</literal> and <literal>dev</literal> outputs.</para>
  </listitem>

</itemizedlist>

</para>

<para>The following builtin functions are new or extended:

<itemizedlist>

  <listitem>
    <para><function xlink:href="https://github.com/NixOS/nix/commit/38539b943a060d9cdfc24d6e5d997c0885b8aa2f">builtins.fetchGit</function>
    allows Git repositories to be fetched at evaluation time. Thus it
    differs from the <function>fetchgit</function> function in
    Nixpkgs, which fetches at build time and cannot be used to fetch
    Nix expressions during evaluation. A typical use case is to import
    external NixOS modules from your configuration, e.g.

    <programlisting>imports = [ (builtins.fetchGit https://github.com/edolstra/dwarffs + "/module.nix") ];</programlisting>

    </para>
  </listitem>

  <listitem>
    <para>Similarly, <function>builtins.fetchMercurial</function>
    allows you to fetch Mercurial repositories.</para>
  </listitem>

  <listitem>
    <para><function>builtins.path</function> generalises
    <function>builtins.filterSource</function> and path literals
    (e.g. <literal>./foo</literal>). It allows specifying a store path
    name that differs from the source path name
    (e.g. <literal>builtins.path { path = ./foo; name = "bar";
    }</literal>) and also supports filtering out unwanted
    files.</para>
  </listitem>

  <listitem>
    <para><function>builtins.fetchurl</function> and
    <function>builtins.fetchTarball</function> now support
    <varname>sha256</varname> and <varname>name</varname>
    attributes.</para>
  </listitem>

  <listitem>
    <para><function xlink:href="https://github.com/NixOS/nix/commit/b8867a0239b1930a16f9ef3f7f3e864b01416dff">builtins.split</function>
    splits a string using a POSIX extended regular expression as the
    separator.</para>
  </listitem>

  <listitem>
    <para><function xlink:href="https://github.com/NixOS/nix/commit/26d92017d3b36cff940dcb7d1611c42232edb81a">builtins.partition</function>
    partitions the elements of a list into two lists, depending on a
    Boolean predicate.</para>
  </listitem>

  <listitem>
    <para><literal>&lt;nix/fetchurl.nix&gt;</literal> now uses the
    content-addressable tarball cache at
    <uri>http://tarballs.nixos.org/</uri>, just like
    <function>fetchurl</function> in
    Nixpkgs. (f2682e6e18a76ecbfb8a12c17e3a0ca15c084197)</para>
  </listitem>

  <listitem>
    <para>In restricted and pure evaluation mode, builtin functions
    that download from the network (such as
    <function>fetchGit</function>) are permitted to fetch underneath a
    list of URI prefixes specified in the option
    <option>allowed-uris</option>.</para>
  </listitem>

</itemizedlist>

</para>

<para>The Nix build environment has the following changes:

<itemizedlist>

  <listitem>
    <para>Values such as Booleans, integers, (nested) lists and
    attribute sets can <link xlink:href="https://github.com/NixOS/nix/commit/6de33a9c675b187437a2e1abbcb290981a89ecb1">now</link>
    be passed to builders in a non-lossy way. If the special attribute
    <varname>__structuredAttrs</varname> is set to
    <literal>true</literal>, the other derivation attributes are
    serialised in JSON format and made available to the builder via
    the file <envar>.attrs.json</envar> in the builder&#x2019;s temporary
    directory. This obviates the need for
    <varname>passAsFile</varname> since JSON files have no size
    restrictions, unlike process environments.</para>

    <para><link xlink:href="https://github.com/NixOS/nix/commit/2d5b1b24bf70a498e4c0b378704cfdb6471cc699">As
    a convenience to Bash builders</link>, Nix writes a script named
    <envar>.attrs.sh</envar> to the builder&#x2019;s directory that
    initialises shell variables corresponding to all attributes that
    are representable in Bash. This includes non-nested (associative)
    arrays. For example, the attribute <literal>hardening.format =
    true</literal> ends up as the Bash associative array element
    <literal>${hardening[format]}</literal>.</para>
  </listitem>

  <listitem>
    <para>Builders can <link xlink:href="https://github.com/NixOS/nix/commit/88e6bb76de5564b3217be9688677d1c89101b2a3">now</link>
    communicate what build phase they are in by writing messages to
    the file descriptor specified in <envar>NIX_LOG_FD</envar>. The
    current phase is shown by the <command>nix</command> progress
    indicator.
    </para>
  </listitem>

  <listitem>
    <para>In Linux sandbox builds, we <link xlink:href="https://github.com/NixOS/nix/commit/a2d92bb20e82a0957067ede60e91fab256948b41">now</link>
    provide a default <filename>/bin/sh</filename> (namely
    <filename>ash</filename> from BusyBox).</para>
  </listitem>

  <listitem>
    <para>In structured attribute mode,
    <varname>exportReferencesGraph</varname> <link xlink:href="https://github.com/NixOS/nix/commit/c2b0d8749f7e77afc1c4b3e8dd36b7ee9720af4a">exports</link>
    extended information about closures in JSON format. In particular,
    it includes the sizes and hashes of paths. This is primarily
    useful for NixOS image builders.</para>
  </listitem>

  <listitem>
    <para>Builds are <link xlink:href="https://github.com/NixOS/nix/commit/21948deed99a3295e4d5666e027a6ca42dc00b40">now</link>
    killed as soon as Nix receives EOF on the builder&#x2019;s stdout or
    stderr. This fixes a bug that allowed builds to hang Nix
    indefinitely, regardless of
    timeouts.</para>
  </listitem>

  <listitem>
    <para>The <option>sandbox-paths</option> configuration
    option can now specify optional paths by appending a
    <literal>?</literal>, e.g. <literal>/dev/nvidiactl?</literal> will
    bind-mount <varname>/dev/nvidiactl</varname> only if it
    exists.</para>
  </listitem>

  <listitem>
    <para>On Linux, builds are now executed in a user
    namespace with UID 1000 and GID 100.</para>
  </listitem>

</itemizedlist>

</para>

<para>A number of significant internal changes were made:

<itemizedlist>

  <listitem>
    <para>Nix no longer depends on Perl and all Perl components have
    been rewritten in C++ or removed. The Perl bindings that used to
    be part of Nix have been moved to a separate package,
    <literal>nix-perl</literal>.</para>
  </listitem>

  <listitem>
    <para>All <classname>Store</classname> classes are now
    thread-safe. <classname>RemoteStore</classname> supports multiple
    concurrent connections to the daemon. This is primarily useful in
    multi-threaded programs such as
    <command>hydra-queue-runner</command>.</para>
  </listitem>

</itemizedlist>

</para>

<para>This release has contributions from

Adrien Devresse,
Alexander Ried,
Alex Cruice,
Alexey Shmalko,
AmineChikhaoui,
Andy Wingo,
Aneesh Agrawal,
Anthony Cowley,
Armijn Hemel,
aszlig,
Ben Gamari,
Benjamin Hipple,
Benjamin Staffin,
Benno F&#xFC;nfst&#xFC;ck,
Bj&#xF8;rn Forsman,
Brian McKenna,
Charles Strahan,
Chase Adams,
Chris Martin,
Christian Theune,
Chris Warburton,
Daiderd Jordan,
Dan Connolly,
Daniel Peebles,
Dan Peebles,
davidak,
David McFarland,
Dmitry Kalinkin,
Domen Ko&#x17E;ar,
Eelco Dolstra,
Emery Hemingway,
Eric Litak,
Eric Wolf,
Fabian Schmitthenner,
Frederik Rietdijk,
Gabriel Gonzalez,
Giorgio Gallo,
Graham Christensen,
Guillaume Maudoux,
Harmen,
Iavael,
James Broadhead,
James Earl Douglas,
Janus Troelsen,
Jeremy Shaw,
Joachim Schiele,
Joe Hermaszewski,
Joel Moberg,
Johannes 'fish' Ziemke,
J&#xF6;rg Thalheim,
Jude Taylor,
kballou,
Keshav Kini,
Kjetil Orbekk,
Langston Barrett,
Linus Heckemann,
Ludovic Court&#xE8;s,
Manav Rathi,
Marc Scholten,
Markus Hauck,
Matt Audesse,
Matthew Bauer,
Matthias Beyer,
Matthieu Coudron,
N1X,
Nathan Zadoks,
Neil Mayhew,
Nicolas B. Pierron,
Niklas Hamb&#xFC;chen,
Nikolay Amiantov,
Ole J&#xF8;rgen Br&#xF8;nner,
Orivej Desh,
Peter Simons,
Peter Stuart,
Pyry Jahkola,
regnat,
Renzo Carbonara,
Rhys,
Robert Vollmert,
Scott Olson,
Scott R. Parish,
Sergei Trofimovich,
Shea Levy,
Sheena Artrip,
Spencer Baugh,
Stefan Junker,
Susan Potter,
Thomas Tuegel,
Timothy Allen,
Tristan Hume,
Tuomas Tynkkynen,
tv,
Tyson Whitehead,
Vladim&#xED;r &#x10C;un&#xE1;t,
Will Dietz,
wmertens,
Wout Mertens,
zimbatm and
Zoran Plesiv&#x10D;ak.
</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-1.11.10">

<title>Release 1.11.10 (2017-06-12)</title>

<para>This release fixes a security bug in Nix&#x2019;s &#x201C;build user&#x201D; build
isolation mechanism. Previously, Nix builders had the ability to
create setuid binaries owned by a <literal>nixbld</literal>
user. Such a binary could then be used by an attacker to assume a
<literal>nixbld</literal> identity and interfere with subsequent
builds running under the same UID.</para>

<para>To prevent this issue, Nix now disallows builders to create
setuid and setgid binaries. On Linux, this is done using a seccomp BPF
filter. Note that this imposes a small performance penalty (e.g. 1%
when building GNU Hello). Using seccomp, we now also prevent the
creation of extended attributes and POSIX ACLs since these cannot be
represented in the NAR format and (in the case of POSIX ACLs) allow
bypassing regular Nix store permissions. On macOS, the restriction is
implemented using the existing sandbox mechanism, which now uses a
minimal &#x201C;allow all except the creation of setuid/setgid binaries&#x201D;
profile when regular sandboxing is disabled. On other platforms, the
&#x201C;build user&#x201D; mechanism is now disabled.</para>

<para>Thanks go to Linus Heckemann for discovering and reporting this
bug.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-1.11">

<title>Release 1.11 (2016-01-19)</title>

<para>This is primarily a bug fix release. It also has a number of new
features:</para>

<itemizedlist>

  <listitem>
    <para><command>nix-prefetch-url</command> can now download URLs
    specified in a Nix expression. For example,

<screen>
$ nix-prefetch-url -A hello.src
</screen>

    will prefetch the file specified by the
    <function>fetchurl</function> call in the attribute
    <literal>hello.src</literal> from the Nix expression in the
    current directory, and print the cryptographic hash of the
    resulting file on stdout. This differs from <literal>nix-build -A
    hello.src</literal> in that it doesn't verify the hash, and is
    thus useful when you&#x2019;re updating a Nix expression.</para>

    <para>You can also prefetch the result of functions that unpack a
    tarball, such as <function>fetchFromGitHub</function>. For example:

<screen>
$ nix-prefetch-url --unpack https://github.com/NixOS/patchelf/archive/0.8.tar.gz
</screen>

    or from a Nix expression:

<screen>
$ nix-prefetch-url -A nix-repl.src
</screen>

    </para>

  </listitem>

  <listitem>
    <para>The builtin function
    <function>&lt;nix/fetchurl.nix&gt;</function> now supports
    downloading and unpacking NARs. This removes the need to have
    multiple downloads in the Nixpkgs stdenv bootstrap process (like a
    separate busybox binary for Linux, or curl/mkdir/sh/bzip2 for
    Darwin). Now all those files can be combined into a single NAR,
    optionally compressed using <command>xz</command>.</para>
  </listitem>

  <listitem>
    <para>Nix now supports SHA-512 hashes for verifying fixed-output
    derivations, and in <function>builtins.hashString</function>.</para>
  </listitem>

  <listitem>
    <para>
      The new flag <option>--option build-repeat
      <replaceable>N</replaceable></option> will cause every build to
      be executed <replaceable>N</replaceable>+1 times. If the build
      output differs between any round, the build is rejected, and the
      output paths are not registered as valid. This is primarily
      useful to verify build determinism. (We already had a
      <option>--check</option> option to repeat a previously succeeded
      build. However, with <option>--check</option>, non-deterministic
      builds are registered in the DB. Preventing that is useful for
      Hydra to ensure that non-deterministic builds don't end up
      getting published to the binary cache.)
    </para>
  </listitem>

  <listitem>
    <para>
      The options <option>--check</option> and <option>--option
      build-repeat <replaceable>N</replaceable></option>, if they
      detect a difference between two runs of the same derivation and
      <option>-K</option> is given, will make the output of the other
      run available under
      <filename><replaceable>store-path</replaceable>-check</filename>. This
      makes it easier to investigate the non-determinism using tools
      like <command>diffoscope</command>, e.g.,

<screen>
$ nix-build pkgs/stdenv/linux -A stage1.pkgs.zlib --check -K
error: derivation &#x2018;/nix/store/l54i8wlw2265&#x2026;-zlib-1.2.8.drv&#x2019; may not
be deterministic: output &#x2018;/nix/store/11a27shh6n2i&#x2026;-zlib-1.2.8&#x2019;
differs from &#x2018;/nix/store/11a27shh6n2i&#x2026;-zlib-1.2.8-check&#x2019;

$ diffoscope /nix/store/11a27shh6n2i&#x2026;-zlib-1.2.8 /nix/store/11a27shh6n2i&#x2026;-zlib-1.2.8-check
&#x2026;
&#x251C;&#x2500;&#x2500; lib/libz.a
&#x2502;   &#x251C;&#x2500;&#x2500; metadata
&#x2502;   &#x2502; @@ -1,15 +1,15 @@
&#x2502;   &#x2502; -rw-r--r-- 30001/30000   3096 Jan 12 15:20 2016 adler32.o
&#x2026;
&#x2502;   &#x2502; +rw-r--r-- 30001/30000   3096 Jan 12 15:28 2016 adler32.o
&#x2026;
</screen>

    </para></listitem>

  <listitem>
    <para>Improved FreeBSD support.</para>
  </listitem>

  <listitem>
    <para><command>nix-env -qa --xml --meta</command> now prints
    license information.</para>
  </listitem>

  <listitem>
    <para>The maximum number of parallel TCP connections that the
    binary cache substituter will use has been decreased from 150 to
    25. This should prevent upsetting some broken NAT routers, and
    also improves performance.</para>
  </listitem>

  <listitem>
    <para>All "chroot"-containing strings got renamed to "sandbox".
      In particular, some Nix options got renamed, but the old names
      are still accepted as lower-priority aliases.
    </para>
  </listitem>

</itemizedlist>

<para>This release has contributions from Anders Claesson, Anthony
Cowley, Bj&#xF8;rn Forsman, Brian McKenna, Danny Wilson, davidak, Eelco Dolstra,
Fabian Schmitthenner, FrankHB, Ilya Novoselov, janus, Jim Garrison, John
Ericson, Jude Taylor, Ludovic Court&#xE8;s, Manuel Jacob, Mathnerd314,
Pascal Wittmann, Peter Simons, Philip Potter, Preston Bennes, Rommel
M. Martinez, Sander van der Burg, Shea Levy, Tim Cuthbertson, Tuomas
Tynkkynen, Utku Demir and Vladim&#xED;r &#x10C;un&#xE1;t.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-1.10">

<title>Release 1.10 (2015-09-03)</title>

<para>This is primarily a bug fix release. It also has a number of new
features:</para>

<itemizedlist>

  <listitem>
    <para>A number of builtin functions have been added to reduce
    Nixpkgs/NixOS evaluation time and memory consumption:
    <function>all</function>,
    <function>any</function>,
    <function>concatStringsSep</function>,
    <function>foldl&#x2019;</function>,
    <function>genList</function>,
    <function>replaceStrings</function>,
    <function>sort</function>.
    </para>
  </listitem>

  <listitem>
    <para>The garbage collector is more robust when the disk is full.</para>
  </listitem>

  <listitem>
    <para>Nix supports a new API for building derivations that doesn&#x2019;t
    require a <literal>.drv</literal> file to be present on disk; it
    only requires an in-memory representation of the derivation. This
    is used by the Hydra continuous build system to make remote builds
    more efficient.</para>
  </listitem>

  <listitem>
    <para>The function <literal>&lt;nix/fetchurl.nix&gt;</literal> now
    uses a <emphasis>builtin</emphasis> builder (i.e. it doesn&#x2019;t
    require starting an external process; the download is performed by
    Nix itself). This ensures that derivation paths don&#x2019;t change when
    Nix is upgraded, and obviates the need for ugly hacks to support
    chroot execution.</para>
  </listitem>

  <listitem>
    <para><option>--version -v</option> now prints some configuration
    information, in particular what compile-time optional features are
    enabled, and the paths of various directories.</para>
  </listitem>

  <listitem>
    <para>Build users have their supplementary groups set correctly.</para>
  </listitem>

</itemizedlist>

<para>This release has contributions from Eelco Dolstra, Guillaume
Maudoux, Iwan Aucamp, Jaka Hudoklin, Kirill Elagin, Ludovic Court&#xE8;s,
Manolis Ragkousis, Nicolas B. Pierron and Shea Levy.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-1.9">

<title>Release 1.9 (2015-06-12)</title>

<para>In addition to the usual bug fixes, this release has the
following new features:</para>

<itemizedlist>

  <listitem>
    <para>Signed binary cache support. You can enable signature
    checking by adding the following to <filename>nix.conf</filename>:

<programlisting>
signed-binary-caches = *
binary-cache-public-keys = cache.nixos.org-1:6NCHdD59X431o0gWypbMrAURkbJ16ZPMQFGspcDShjY=
</programlisting>

    This will prevent Nix from downloading any binary from the cache
    that is not signed by one of the keys listed in
    <option>binary-cache-public-keys</option>.</para>

    <para>Signature checking is only supported if you built Nix with
    the <literal>libsodium</literal> package.</para>

    <para>Note that while Nix has had experimental support for signed
    binary caches since version 1.7, this release changes the
    signature format in a backwards-incompatible way.</para>

  </listitem>

  <listitem>

    <para>Automatic downloading of Nix expression tarballs. In various
    places, you can now specify the URL of a tarball containing Nix
    expressions (such as Nixpkgs), which will be downloaded and
    unpacked automatically. For example:</para>

    <itemizedlist>

      <listitem><para>In <command>nix-env</command>:

<screen>
$ nix-env -f https://github.com/NixOS/nixpkgs-channels/archive/nixos-14.12.tar.gz -iA firefox
</screen>

      This installs Firefox from the latest tested and built revision
      of the NixOS 14.12 channel.</para></listitem>

      <listitem><para>In <command>nix-build</command> and
      <command>nix-shell</command>:

<screen>
$ nix-build https://github.com/NixOS/nixpkgs/archive/master.tar.gz -A hello
</screen>

      This builds GNU Hello from the latest revision of the Nixpkgs
      master branch.</para></listitem>

      <listitem><para>In the Nix search path (as specified via
      <envar>NIX_PATH</envar> or <option>-I</option>). For example, to
      start a shell containing the Pan package from a specific version
      of Nixpkgs:

<screen>
$ nix-shell -p pan -I nixpkgs=https://github.com/NixOS/nixpkgs-channels/archive/8a3eea054838b55aca962c3fbde9c83c102b8bf2.tar.gz
</screen>

      </para></listitem>

      <listitem><para>In <command>nixos-rebuild</command> (on NixOS):

<screen>
$ nixos-rebuild test -I nixpkgs=https://github.com/NixOS/nixpkgs-channels/archive/nixos-unstable.tar.gz
</screen>

      </para></listitem>

      <listitem><para>In Nix expressions, via the new builtin function <function>fetchTarball</function>:

<programlisting>
with import (fetchTarball https://github.com/NixOS/nixpkgs-channels/archive/nixos-14.12.tar.gz) {}; &#x2026;
</programlisting>

      (This is not allowed in restricted mode.)</para></listitem>

    </itemizedlist>

  </listitem>

  <listitem>

    <para><command>nix-shell</command> improvements:</para>

    <itemizedlist>

      <listitem><para><command>nix-shell</command> now has a flag
      <option>--run</option> to execute a command in the
      <command>nix-shell</command> environment,
      e.g. <literal>nix-shell --run make</literal>. This is like
      the existing <option>--command</option> flag, except that it
      uses a non-interactive shell (ensuring that hitting Ctrl-C won&#x2019;t
      drop you into the child shell).</para></listitem>

      <listitem><para><command>nix-shell</command> can now be used as
      a <literal>#!</literal>-interpreter. This allows you to write
      scripts that dynamically fetch their own dependencies. For
      example, here is a Haskell script that, when invoked, first
      downloads GHC and the Haskell packages on which it depends:

<programlisting>
#! /usr/bin/env nix-shell
#! nix-shell -i runghc -p haskellPackages.ghc haskellPackages.HTTP

import Network.HTTP

main = do
  resp &lt;- Network.HTTP.simpleHTTP (getRequest "http://nixos.org/")
  body &lt;- getResponseBody resp
  print (take 100 body)
</programlisting>

      Of course, the dependencies are cached in the Nix store, so the
      second invocation of this script will be much
      faster.</para></listitem>

    </itemizedlist>

  </listitem>

  <listitem>

    <para>Chroot improvements:</para>

    <itemizedlist>

      <listitem><para>Chroot builds are now supported on Mac OS X
      (using its sandbox mechanism).</para></listitem>

      <listitem><para>If chroots are enabled, they are now used for
      all derivations, including fixed-output derivations (such as
      <function>fetchurl</function>). The latter do have network
      access, but can no longer access the host filesystem. If you
      need the old behaviour, you can set the option
      <option>build-use-chroot</option> to
      <literal>relaxed</literal>.</para></listitem>

      <listitem><para>On Linux, if chroots are enabled, builds are
      performed in a private PID namespace once again. (This
      functionality was lost in Nix 1.8.)</para></listitem>

      <listitem><para>Store paths listed in
      <option>build-chroot-dirs</option> are now automatically
      expanded to their closure. For instance, if you want
      <filename>/nix/store/&#x2026;-bash/bin/sh</filename> mounted in your
      chroot as <filename>/bin/sh</filename>, you only need to say
      <literal>build-chroot-dirs =
      /bin/sh=/nix/store/&#x2026;-bash/bin/sh</literal>; it is no longer
      necessary to specify the dependencies of Bash.</para></listitem>

    </itemizedlist>

  </listitem>

  <listitem><para>The new derivation attribute
  <varname>passAsFile</varname> allows you to specify that the
  contents of derivation attributes should be passed via files rather
  than environment variables. This is useful if you need to pass very
  long strings that exceed the size limit of the environment. The
  Nixpkgs function <function>writeTextFile</function> uses
  this.</para></listitem>

  <listitem><para>You can now use <literal>~</literal> in Nix file
  names to refer to your home directory, e.g. <literal>import
  ~/.nixpkgs/config.nix</literal>.</para></listitem>

  <listitem><para>Nix has a new option <option>restrict-eval</option>
  that allows limiting what paths the Nix evaluator has access to. By
  passing <literal>--option restrict-eval true</literal> to Nix, the
  evaluator will throw an exception if an attempt is made to access
  any file outside of the Nix search path. This is primarily intended
  for Hydra to ensure that a Hydra jobset only refers to its declared
  inputs (and is therefore reproducible).</para></listitem>

  <listitem><para><command>nix-env</command> now only creates a new
  &#x201C;generation&#x201D; symlink in <filename>/nix/var/nix/profiles</filename>
  if something actually changed.</para></listitem>

  <listitem><para>The environment variable <envar>NIX_PAGER</envar>
  can now be set to override <envar>PAGER</envar>. You can set it to
  <literal>cat</literal> to disable paging for Nix commands
  only.</para></listitem>

  <listitem><para>Failing <literal>&lt;...&gt;</literal>
  lookups now show position information.</para></listitem>

  <listitem><para>Improved Boehm GC use: we disabled scanning for
  interior pointers, which should reduce the &#x201C;<literal>Repeated
  allocation of very large block</literal>&#x201D; warnings and associated
  retention of memory.</para></listitem>

</itemizedlist>

<para>This release has contributions from aszlig, Benjamin Staffin,
Charles Strahan, Christian Theune, Daniel Hahler, Danylo Hlynskyi
Daniel Peebles, Dan Peebles, Domen Ko&#x17E;ar, Eelco Dolstra, Harald van
Dijk, Hoang Xuan Phu, Jaka Hudoklin, Jeff Ramnani, j-keck, Linquize,
Luca Bruno, Michael Merickel, Oliver Dunkl, Rob Vermaas, Rok Garbas,
Shea Levy, Tobias Geerinckx-Rice and William A. Kennington III.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-1.8">

<title>Release 1.8 (2014-12-14)</title>

<itemizedlist>

  <listitem><para>Breaking change: to address a race condition, the
  remote build hook mechanism now uses <command>nix-store
  --serve</command> on the remote machine. This requires build slaves
  to be updated to Nix 1.8.</para></listitem>

  <listitem><para>Nix now uses HTTPS instead of HTTP to access the
  default binary cache,
  <literal>cache.nixos.org</literal>.</para></listitem>

  <listitem><para><command>nix-env</command> selectors are now regular
  expressions. For instance, you can do

<screen>
$ nix-env -qa '.*zip.*'
</screen>

  to query all packages with a name containing
  <literal>zip</literal>.</para></listitem>

  <listitem><para><command>nix-store --read-log</command> can now
  fetch remote build logs. If a build log is not available locally,
  then &#x2018;nix-store -l&#x2019; will now try to download it from the servers
  listed in the &#x2018;log-servers&#x2019; option in nix.conf. For instance, if you
  have the configuration option

<programlisting>
log-servers = http://hydra.nixos.org/log
</programlisting>

then it will try to get logs from
<literal>http://hydra.nixos.org/log/<replaceable>base name of the
store path</replaceable></literal>. This allows you to do things like:

<screen>
$ nix-store -l $(which xterm)
</screen>

  and get a log even if <command>xterm</command> wasn't built
  locally.</para></listitem>

  <listitem><para>New builtin functions:
  <function>attrValues</function>, <function>deepSeq</function>,
  <function>fromJSON</function>, <function>readDir</function>,
  <function>seq</function>.</para></listitem>

  <listitem><para><command>nix-instantiate --eval</command> now has a
  <option>--json</option> flag to print the resulting value in JSON
  format.</para></listitem>

  <listitem><para><command>nix-copy-closure</command> now uses
  <command>nix-store --serve</command> on the remote side to send or
  receive closures. This fixes a race condition between
  <command>nix-copy-closure</command> and the garbage
  collector.</para></listitem>

  <listitem><para>Derivations can specify the new special attribute
  <varname>allowedRequisites</varname>, which has a similar meaning to
  <varname>allowedReferences</varname>. But instead of only enforcing
  to explicitly specify the immediate references, it requires the
  derivation to specify all the dependencies recursively (hence the
  name, requisites) that are used by the resulting
  output.</para></listitem>

  <listitem><para>On Mac OS X, Nix now handles case collisions when
  importing closures from case-sensitive file systems. This is mostly
  useful for running NixOps on Mac OS X.</para></listitem>

  <listitem><para>The Nix daemon has new configuration options
  <option>allowed-users</option> (specifying the users and groups that
  are allowed to connect to the daemon) and
  <option>trusted-users</option> (specifying the users and groups that
  can perform privileged operations like specifying untrusted binary
  caches).</para></listitem>

  <listitem><para>The configuration option
  <option>build-cores</option> now defaults to the number of available
  CPU cores.</para></listitem>

  <listitem><para>Build users are now used by default when Nix is
  invoked as root. This prevents builds from accidentally running as
  root.</para></listitem>

  <listitem><para>Nix now includes systemd units and Upstart
  jobs.</para></listitem>

  <listitem><para>Speed improvements to <command>nix-store
  --optimise</command>.</para></listitem>

  <listitem><para>Language change: the <literal>==</literal> operator
  now ignores string contexts (the &#x201C;dependencies&#x201D; of a
  string).</para></listitem>

  <listitem><para>Nix now filters out Nix-specific ANSI escape
  sequences on standard error. They are supposed to be invisible, but
  some terminals show them anyway.</para></listitem>

  <listitem><para>Various commands now automatically pipe their output
  into the pager as specified by the <envar>PAGER</envar> environment
  variable.</para></listitem>

  <listitem><para>Several improvements to reduce memory consumption in
  the evaluator.</para></listitem>

</itemizedlist>

<para>This release has contributions from Adam Szkoda, Aristid
Breitkreuz, Bob van der Linden, Charles Strahan, darealshinji, Eelco
Dolstra, Gergely Risko, Joel Taylor, Ludovic Court&#xE8;s, Marko Durkovic,
Mikey Ariel, Paul Colomiets, Ricardo M.  Correia, Ricky Elrod, Robert
Helgesson, Rob Vermaas, Russell O'Connor, Shea Levy, Shell Turner,
S&#xF6;nke Hahn, Steve Purcell, Vladim&#xED;r &#x10C;un&#xE1;t and Wout Mertens.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-1.7">

<title>Release 1.7 (2014-04-11)</title>

<para>In addition to the usual bug fixes, this release has the
following new features:</para>

<itemizedlist>

  <listitem>
    <para>Antiquotation is now allowed inside of quoted attribute
    names (e.g. <literal>set."${foo}"</literal>). In the case where
    the attribute name is just a single antiquotation, the quotes can
    be dropped (e.g. the above example can be written
    <literal>set.${foo}</literal>). If an attribute name inside of a
    set declaration evaluates to <literal>null</literal> (e.g.
    <literal>{ ${null} = false; }</literal>), then that attribute is
    not added to the set.</para>
  </listitem>

  <listitem>
    <para>Experimental support for cryptographically signed binary
    caches.  See <link xlink:href="https://github.com/NixOS/nix/commit/0fdf4da0e979f992db75cc17376e455ddc5a96d8">the
    commit for details</link>.</para>
  </listitem>

  <listitem>
    <para>An experimental new substituter,
    <command>download-via-ssh</command>, that fetches binaries from
    remote machines via SSH.  Specifying the flags <literal>--option
    use-ssh-substituter true --option ssh-substituter-hosts
    <replaceable>user@hostname</replaceable></literal> will cause Nix
    to download binaries from the specified machine, if it has
    them.</para>
  </listitem>

  <listitem>
    <para><command>nix-store -r</command> and
    <command>nix-build</command> have a new flag,
    <option>--check</option>, that builds a previously built
    derivation again, and prints an error message if the output is not
    exactly the same. This helps to verify whether a derivation is
    truly deterministic.  For example:

<screen>
$ nix-build '&lt;nixpkgs&gt;' -A patchelf
<replaceable>&#x2026;</replaceable>
$ nix-build '&lt;nixpkgs&gt;' -A patchelf --check
<replaceable>&#x2026;</replaceable>
error: derivation `/nix/store/1ipvxs&#x2026;-patchelf-0.6' may not be deterministic:
  hash mismatch in output `/nix/store/4pc1dm&#x2026;-patchelf-0.6.drv'
</screen>

    </para>

  </listitem>

  <listitem>
    <para>The <command>nix-instantiate</command> flags
    <option>--eval-only</option> and <option>--parse-only</option>
    have been renamed to <option>--eval</option> and
    <option>--parse</option>, respectively.</para>
  </listitem>

  <listitem>
    <para><command>nix-instantiate</command>,
    <command>nix-build</command> and <command>nix-shell</command> now
    have a flag <option>--expr</option> (or <option>-E</option>) that
    allows you to specify the expression to be evaluated as a command
    line argument.  For instance, <literal>nix-instantiate --eval -E
    '1 + 2'</literal> will print <literal>3</literal>.</para>
  </listitem>

  <listitem>
    <para><command>nix-shell</command> improvements:</para>

    <itemizedlist>

      <listitem>
        <para>It has a new flag, <option>--packages</option> (or
        <option>-p</option>), that sets up a build environment
        containing the specified packages from Nixpkgs. For example,
        the command

<screen>
$ nix-shell -p sqlite xorg.libX11 hello
</screen>

        will start a shell in which the given packages are
        present.</para>
      </listitem>

      <listitem>
        <para>It now uses <filename>shell.nix</filename> as the
        default expression, falling back to
        <filename>default.nix</filename> if the former doesn&#x2019;t
        exist.  This makes it convenient to have a
        <filename>shell.nix</filename> in your project to set up a
        nice development environment.</para>
      </listitem>

      <listitem>
        <para>It evaluates the derivation attribute
        <varname>shellHook</varname>, if set. Since
        <literal>stdenv</literal> does not normally execute this hook,
        it allows you to do <command>nix-shell</command>-specific
        setup.</para>
      </listitem>

      <listitem>
        <para>It preserves the user&#x2019;s timezone setting.</para>
      </listitem>

    </itemizedlist>

  </listitem>

  <listitem>
    <para>In chroots, Nix now sets up a <filename>/dev</filename>
    containing only a minimal set of devices (such as
    <filename>/dev/null</filename>). Note that it only does this if
    you <emphasis>don&#x2019;t</emphasis> have <filename>/dev</filename>
    listed in your <option>build-chroot-dirs</option> setting;
    otherwise, it will bind-mount the <literal>/dev</literal> from
    outside the chroot.</para>

    <para>Similarly, if you don&#x2019;t have <filename>/dev/pts</filename> listed
    in <option>build-chroot-dirs</option>, Nix will mount a private
    <literal>devpts</literal> filesystem on the chroot&#x2019;s
    <filename>/dev/pts</filename>.</para>

  </listitem>

  <listitem>
    <para>New built-in function: <function>builtins.toJSON</function>,
    which returns a JSON representation of a value.</para>
  </listitem>

  <listitem>
    <para><command>nix-env -q</command> has a new flag
    <option>--json</option> to print a JSON representation of the
    installed or available packages.</para>
  </listitem>

  <listitem>
    <para><command>nix-env</command> now supports meta attributes with
    more complex values, such as attribute sets.</para>
  </listitem>

  <listitem>
    <para>The <option>-A</option> flag now allows attribute names with
    dots in them, e.g.

<screen>
$ nix-instantiate --eval '&lt;nixos&gt;' -A 'config.systemd.units."nscd.service".text'
</screen>

    </para>
  </listitem>

  <listitem>
    <para>The <option>--max-freed</option> option to
    <command>nix-store --gc</command> now accepts a unit
    specifier. For example, <literal>nix-store --gc --max-freed
    1G</literal> will free up to 1 gigabyte of disk space.</para>
  </listitem>

  <listitem>
    <para><command>nix-collect-garbage</command> has a new flag
    <option>--delete-older-than</option>
    <replaceable>N</replaceable><literal>d</literal>, which deletes
    all user environment generations older than
    <replaceable>N</replaceable> days.  Likewise, <command>nix-env
    --delete-generations</command> accepts a
    <replaceable>N</replaceable><literal>d</literal> age limit.</para>
  </listitem>

  <listitem>
    <para>Nix now heuristically detects whether a build failure was
    due to a disk-full condition. In that case, the build is not
    flagged as &#x201C;permanently failed&#x201D;. This is mostly useful for Hydra,
    which needs to distinguish between permanent and transient build
    failures.</para>
  </listitem>

  <listitem>
    <para>There is a new symbol <literal>__curPos</literal> that
    expands to an attribute set containing its file name and line and
    column numbers, e.g. <literal>{ file = "foo.nix"; line = 10;
    column = 5; }</literal>.  There also is a new builtin function,
    <varname>unsafeGetAttrPos</varname>, that returns the position of
    an attribute.  This is used by Nixpkgs to provide location
    information in error messages, e.g.

<screen>
$ nix-build '&lt;nixpkgs&gt;' -A libreoffice --argstr system x86_64-darwin
error: the package &#x2018;libreoffice-4.0.5.2&#x2019; in &#x2018;.../applications/office/libreoffice/default.nix:263&#x2019;
  is not supported on &#x2018;x86_64-darwin&#x2019;
</screen>

    </para>
  </listitem>

  <listitem>
    <para>The garbage collector is now more concurrent with other Nix
    processes because it releases certain locks earlier.</para>
  </listitem>

  <listitem>
    <para>The binary tarball installer has been improved.  You can now
    install Nix by running:

<screen>
$ bash &lt;(curl https://nixos.org/nix/install)
</screen>

    </para>
  </listitem>

  <listitem>
    <para>More evaluation errors include position information. For
    instance, selecting a missing attribute will print something like

<screen>
error: attribute `nixUnstabl' missing, at /etc/nixos/configurations/misc/eelco/mandark.nix:216:15
</screen>

    </para>
  </listitem>

  <listitem>
    <para>The command <command>nix-setuid-helper</command> is
    gone.</para>
  </listitem>

  <listitem>
    <para>Nix no longer uses Automake, but instead has a
    non-recursive, GNU Make-based build system.</para>
  </listitem>

  <listitem>
    <para>All installed libraries now have the prefix
    <literal>libnix</literal>.  In particular, this gets rid of
    <literal>libutil</literal>, which could clash with libraries with
    the same name from other packages.</para>
  </listitem>

  <listitem>
    <para>Nix now requires a compiler that supports C++11.</para>
  </listitem>

</itemizedlist>

<para>This release has contributions from Danny Wilson, Domen Ko&#x17E;ar,
Eelco Dolstra, Ian-Woo Kim, Ludovic Court&#xE8;s, Maxim Ivanov, Petr
Rockai, Ricardo M. Correia and Shea Levy.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-1.6.1">

<title>Release 1.6.1 (2013-10-28)</title>

<para>This is primarily a bug fix release.  Changes of interest
are:</para>

<itemizedlist>

  <listitem>
    <para>Nix 1.6 accidentally changed the semantics of antiquoted
    paths in strings, such as <literal>"${/foo}/bar"</literal>.  This
    release reverts to the Nix 1.5.3 behaviour.</para>
  </listitem>

  <listitem>
    <para>Previously, Nix optimised expressions such as
    <literal>"${<replaceable>expr</replaceable>}"</literal> to
    <replaceable>expr</replaceable>.  Thus it neither checked whether
    <replaceable>expr</replaceable> could be coerced to a string, nor
    applied such coercions.  This meant that
    <literal>"${123}"</literal> evaluatued to <literal>123</literal>,
    and <literal>"${./foo}"</literal> evaluated to
    <literal>./foo</literal> (even though
    <literal>"${./foo} "</literal> evaluates to
    <literal>"/nix/store/<replaceable>hash</replaceable>-foo "</literal>).
    Nix now checks the type of antiquoted expressions and
    applies coercions.</para>
  </listitem>

  <listitem>
    <para>Nix now shows the exact position of undefined variables.  In
    particular, undefined variable errors in a <literal>with</literal>
    previously didn't show <emphasis>any</emphasis> position
    information, so this makes it a lot easier to fix such
    errors.</para>
  </listitem>

  <listitem>
    <para>Undefined variables are now treated consistently.
    Previously, the <function>tryEval</function> function would catch
    undefined variables inside a <literal>with</literal> but not
    outside.  Now <function>tryEval</function> never catches undefined
    variables.</para>
  </listitem>

  <listitem>
    <para>Bash completion in <command>nix-shell</command> now works
    correctly.</para>
  </listitem>

  <listitem>
    <para>Stack traces are less verbose: they no longer show calls to
    builtin functions and only show a single line for each derivation
    on the call stack.</para>
  </listitem>

  <listitem>
    <para>New built-in function: <function>builtins.typeOf</function>,
    which returns the type of its argument as a string.</para>
  </listitem>

</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-1.6.0">

<title>Release 1.6 (2013-09-10)</title>

<para>In addition to the usual bug fixes, this release has several new
features:</para>

<itemizedlist>

  <listitem>
    <para>The command <command>nix-build --run-env</command> has been
    renamed to <command>nix-shell</command>.</para>
  </listitem>

  <listitem>
    <para><command>nix-shell</command> now sources
    <filename>$stdenv/setup</filename> <emphasis>inside</emphasis> the
    interactive shell, rather than in a parent shell.  This ensures
    that shell functions defined by <literal>stdenv</literal> can be
    used in the interactive shell.</para>
  </listitem>

  <listitem>
    <para><command>nix-shell</command> has a new flag
    <option>--pure</option> to clear the environment, so you get an
    environment that more closely corresponds to the &#x201C;real&#x201D; Nix build.
    </para>
  </listitem>

  <listitem>
    <para><command>nix-shell</command> now sets the shell prompt
    (<envar>PS1</envar>) to ensure that Nix shells are distinguishable
    from your regular shells.</para>
  </listitem>

  <listitem>
    <para><command>nix-env</command> no longer requires a
    <literal>*</literal> argument to match all packages, so
    <literal>nix-env -qa</literal> is equivalent to <literal>nix-env
    -qa '*'</literal>.</para>
  </listitem>

  <listitem>
    <para><command>nix-env -i</command> has a new flag
    <option>--remove-all</option> (<option>-r</option>) to remove all
    previous packages from the profile.  This makes it easier to do
    declarative package management similar to NixOS&#x2019;s
    <option>environment.systemPackages</option>.  For instance, if you
    have a specification <filename>my-packages.nix</filename> like this:

<programlisting>
with import &lt;nixpkgs&gt; {};
[ thunderbird
  geeqie
  ...
]
</programlisting>

    then after any change to this file, you can run:

<screen>
$ nix-env -f my-packages.nix -ir
</screen>

    to update your profile to match the specification.</para>
  </listitem>

  <listitem>
    <para>The &#x2018;<literal>with</literal>&#x2019; language construct is now more
    lazy.  It only evaluates its argument if a variable might actually
    refer to an attribute in the argument.  For instance, this now
    works:

<programlisting>
let
  pkgs = with pkgs; { foo = "old"; bar = foo; } // overrides;
  overrides = { foo = "new"; };
in pkgs.bar
</programlisting>

    This evaluates to <literal>"new"</literal>, while previously it
    gave an &#x201C;infinite recursion&#x201D; error.</para>
  </listitem>

  <listitem>
    <para>Nix now has proper integer arithmetic operators. For
    instance, you can write <literal>x + y</literal> instead of
    <literal>builtins.add x y</literal>, or <literal>x &lt;
    y</literal> instead of <literal>builtins.lessThan x y</literal>.
    The comparison operators also work on strings.</para>
  </listitem>

  <listitem>
    <para>On 64-bit systems, Nix integers are now 64 bits rather than
    32 bits.</para>
  </listitem>

  <listitem>
    <para>When using the Nix daemon, the <command>nix-daemon</command>
    worker process now runs on the same CPU as the client, on systems
    that support setting CPU affinity.  This gives a significant speedup
    on some systems.</para>
  </listitem>

  <listitem>
    <para>If a stack overflow occurs in the Nix evaluator, you now get
    a proper error message (rather than &#x201C;Segmentation fault&#x201D;) on some
    systems.</para>
  </listitem>

  <listitem>
    <para>In addition to directories, you can now bind-mount regular
    files in chroots through the (now misnamed) option
    <option>build-chroot-dirs</option>.</para>
  </listitem>

</itemizedlist>

<para>This release has contributions from Domen Ko&#x17E;ar, Eelco Dolstra,
Florian Friesdorf, Gergely Risko, Ivan Kozik, Ludovic Court&#xE8;s and Shea
Levy.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-1.5.2">

<title>Release 1.5.2 (2013-05-13)</title>

<para>This is primarily a bug fix release.  It has contributions from
Eelco Dolstra, Llu&#xED;s Batlle i Rossell and Shea Levy.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-1.5">

<title>Release 1.5 (2013-02-27)</title>

<para>This is a brown paper bag release to fix a regression introduced
by the hard link security fix in 1.4.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-1.4">

<title>Release 1.4 (2013-02-26)</title>

<para>This release fixes a security bug in multi-user operation.  It
was possible for derivations to cause the mode of files outside of the
Nix store to be changed to 444 (read-only but world-readable) by
creating hard links to those files (<link xlink:href="https://github.com/NixOS/nix/commit/5526a282b5b44e9296e61e07d7d2626a79141ac4">details</link>).</para>

<para>There are also the following improvements:</para>

<itemizedlist>

  <listitem><para>New built-in function:
  <function>builtins.hashString</function>.</para></listitem>

  <listitem><para>Build logs are now stored in
  <filename>/nix/var/log/nix/drvs/<replaceable>XX</replaceable>/</filename>,
  where <replaceable>XX</replaceable> is the first two characters of
  the derivation.  This is useful on machines that keep a lot of build
  logs (such as Hydra servers).</para></listitem>

  <listitem><para>The function <function>corepkgs/fetchurl</function>
  can now make the downloaded file executable.  This will allow
  getting rid of all bootstrap binaries in the Nixpkgs source
  tree.</para></listitem>

  <listitem><para>Language change: The expression <literal>"${./path}
  ..."</literal> now evaluates to a string instead of a
  path.</para></listitem>

</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-1.3">

<title>Release 1.3 (2013-01-04)</title>

<para>This is primarily a bug fix release.  When this version is first
run on Linux, it removes any immutable bits from the Nix store and
increases the schema version of the Nix store.  (The previous release
removed support for setting the immutable bit; this release clears any
remaining immutable bits to make certain operations more
efficient.)</para>

<para>This release has contributions from Eelco Dolstra and Stuart
Pernsteiner.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-1.2">

<title>Release 1.2 (2012-12-06)</title>

<para>This release has the following improvements and changes:</para>

<itemizedlist>

  <listitem>
    <para>Nix has a new binary substituter mechanism: the
    <emphasis>binary cache</emphasis>.  A binary cache contains
    pre-built binaries of Nix packages.  Whenever Nix wants to build a
    missing Nix store path, it will check a set of binary caches to
    see if any of them has a pre-built binary of that path.  The
    configuration setting <option>binary-caches</option> contains a
    list of URLs of binary caches.  For instance, doing
<screen>
$ nix-env -i thunderbird --option binary-caches http://cache.nixos.org
</screen>
    will install Thunderbird and its dependencies, using the available
    pre-built binaries in <uri>http://cache.nixos.org</uri>.
    The main advantage over the old &#x201C;manifest&#x201D;-based method of getting
    pre-built binaries is that you don&#x2019;t have to worry about your
    manifest being in sync with the Nix expressions you&#x2019;re installing
    from; i.e., you don&#x2019;t need to run <command>nix-pull</command> to
    update your manifest.  It&#x2019;s also more scalable because you don&#x2019;t
    need to redownload a giant manifest file every time.
    </para>

    <para>A Nix channel can provide a binary cache URL that will be
    used automatically if you subscribe to that channel.  If you use
    the Nixpkgs or NixOS channels
    (<uri>http://nixos.org/channels</uri>) you automatically get the
    cache <uri>http://cache.nixos.org</uri>.</para>

    <para>Binary caches are created using <command>nix-push</command>.
    For details on the operation and format of binary caches, see the
    <command>nix-push</command> manpage.  More details are provided in
    <link xlink:href="https://nixos.org/nix-dev/2012-September/009826.html">this
    nix-dev posting</link>.</para>
  </listitem>

  <listitem>
    <para>Multiple output support should now be usable.  A derivation
    can declare that it wants to produce multiple store paths by
    saying something like
<programlisting>
outputs = [ "lib" "headers" "doc" ];
</programlisting>
    This will cause Nix to pass the intended store path of each output
    to the builder through the environment variables
    <literal>lib</literal>, <literal>headers</literal> and
    <literal>doc</literal>.  Other packages can refer to a specific
    output by referring to
    <literal><replaceable>pkg</replaceable>.<replaceable>output</replaceable></literal>,
    e.g.
<programlisting>
buildInputs = [ pkg.lib pkg.headers ];
</programlisting>
    If you install a package with multiple outputs using
    <command>nix-env</command>, each output path will be symlinked
    into the user environment.</para>
  </listitem>

  <listitem>
    <para>Dashes are now valid as part of identifiers and attribute
    names.</para>
  </listitem>

  <listitem>
    <para>The new operation <command>nix-store --repair-path</command>
    allows corrupted or missing store paths to be repaired by
    redownloading them.  <command>nix-store --verify --check-contents
    --repair</command> will scan and repair all paths in the Nix
    store.  Similarly, <command>nix-env</command>,
    <command>nix-build</command>, <command>nix-instantiate</command>
    and <command>nix-store --realise</command> have a
    <option>--repair</option> flag to detect and fix bad paths by
    rebuilding or redownloading them.</para>
  </listitem>

  <listitem>
    <para>Nix no longer sets the immutable bit on files in the Nix
    store.  Instead, the recommended way to guard the Nix store
    against accidental modification on Linux is to make it a read-only
    bind mount, like this:

<screen>
$ mount --bind /nix/store /nix/store
$ mount -o remount,ro,bind /nix/store
</screen>

    Nix will automatically make <filename>/nix/store</filename>
    writable as needed (using a private mount namespace) to allow
    modifications.</para>
  </listitem>

  <listitem>
    <para>Store optimisation (replacing identical files in the store
    with hard links) can now be done automatically every time a path
    is added to the store.  This is enabled by setting the
    configuration option <literal>auto-optimise-store</literal> to
    <literal>true</literal> (disabled by default).</para>
  </listitem>

  <listitem>
    <para>Nix now supports <command>xz</command> compression for NARs
    in addition to <command>bzip2</command>.  It compresses about 30%
    better on typical archives and decompresses about twice as
    fast.</para>
  </listitem>

  <listitem>
    <para>Basic Nix expression evaluation profiling: setting the
    environment variable <envar>NIX_COUNT_CALLS</envar> to
    <literal>1</literal> will cause Nix to print how many times each
    primop or function was executed.</para>
  </listitem>

  <listitem>
    <para>New primops: <varname>concatLists</varname>,
    <varname>elem</varname>, <varname>elemAt</varname> and
    <varname>filter</varname>.</para>
  </listitem>

  <listitem>
    <para>The command <command>nix-copy-closure</command> has a new
    flag <option>--use-substitutes</option> (<option>-s</option>) to
    download missing paths on the target machine using the substitute
    mechanism.</para>
  </listitem>

  <listitem>
    <para>The command <command>nix-worker</command> has been renamed
    to <command>nix-daemon</command>.  Support for running the Nix
    worker in &#x201C;slave&#x201D; mode has been removed.</para>
  </listitem>

  <listitem>
    <para>The <option>--help</option> flag of every Nix command now
    invokes <command>man</command>.</para>
  </listitem>

  <listitem>
    <para>Chroot builds are now supported on systemd machines.</para>
  </listitem>

</itemizedlist>

<para>This release has contributions from Eelco Dolstra, Florian
Friesdorf, Mats Erik Andersson and Shea Levy.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-1.1">

<title>Release 1.1 (2012-07-18)</title>

<para>This release has the following improvements:</para>

<itemizedlist>

  <listitem>
    <para>On Linux, when doing a chroot build, Nix now uses various
    namespace features provided by the Linux kernel to improve
    build isolation.  Namely:
    <itemizedlist>
      <listitem><para>The private network namespace ensures that
      builders cannot talk to the outside world (or vice versa): each
      build only sees a private loopback interface.  This also means
      that two concurrent builds can listen on the same port (e.g. as
      part of a test) without conflicting with each
      other.</para></listitem>
      <listitem><para>The PID namespace causes each build to start as
      PID 1.  Processes outside of the chroot are not visible to those
      on the inside.  On the other hand, processes inside the chroot
      <emphasis>are</emphasis> visible from the outside (though with
      different PIDs).</para></listitem>
      <listitem><para>The IPC namespace prevents the builder from
      communicating with outside processes using SysV IPC mechanisms
      (shared memory, message queues, semaphores).  It also ensures
      that all IPC objects are destroyed when the builder
      exits.</para></listitem>
      <listitem><para>The UTS namespace ensures that builders see a
      hostname of <literal>localhost</literal> rather than the actual
      hostname.</para></listitem>
      <listitem><para>The private mount namespace was already used by
      Nix to ensure that the bind-mounts used to set up the chroot are
      cleaned up automatically.</para></listitem>
    </itemizedlist>
    </para>
  </listitem>

  <listitem>
    <para>Build logs are now compressed using
    <command>bzip2</command>.  The command <command>nix-store
    -l</command> decompresses them on the fly.  This can be disabled
    by setting the option <literal>build-compress-log</literal> to
    <literal>false</literal>.</para>
  </listitem>

  <listitem>
    <para>The creation of build logs in
    <filename>/nix/var/log/nix/drvs</filename> can be disabled by
    setting the new option <literal>build-keep-log</literal> to
    <literal>false</literal>.  This is useful, for instance, for Hydra
    build machines.</para>
  </listitem>

  <listitem>
    <para>Nix now reserves some space in
    <filename>/nix/var/nix/db/reserved</filename> to ensure that the
    garbage collector can run successfully if the disk is full.  This
    is necessary because SQLite transactions fail if the disk is
    full.</para>
  </listitem>

  <listitem>
    <para>Added a basic <function>fetchurl</function> function.  This
    is not intended to replace the <function>fetchurl</function> in
    Nixpkgs, but is useful for bootstrapping; e.g., it will allow us
    to get rid of the bootstrap binaries in the Nixpkgs source tree
    and download them instead.  You can use it by doing
    <literal>import &lt;nix/fetchurl.nix&gt; { url =
    <replaceable>url</replaceable>; sha256 =
    "<replaceable>hash</replaceable>"; }</literal>. (Shea Levy)</para>
  </listitem>

  <listitem>
    <para>Improved RPM spec file. (Michel Alexandre Salim)</para>
  </listitem>

  <listitem>
    <para>Support for on-demand socket-based activation in the Nix
    daemon with <command>systemd</command>.</para>
  </listitem>

  <listitem>
    <para>Added a manpage for
    <citerefentry><refentrytitle>nix.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>.</para>
  </listitem>

  <listitem>
    <para>When using the Nix daemon, the <option>-s</option> flag in
    <command>nix-env -qa</command> is now much faster.</para>
  </listitem>

</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-1.0">

<title>Release 1.0 (2012-05-11)</title>

<para>There have been numerous improvements and bug fixes since the
previous release.  Here are the most significant:</para>

<itemizedlist>

  <listitem>
    <para>Nix can now optionally use the Boehm garbage collector.
    This significantly reduces the Nix evaluator&#x2019;s memory footprint,
    especially when evaluating large NixOS system configurations.  It
    can be enabled using the <option>--enable-gc</option> configure
    option.</para>
  </listitem>

  <listitem>
    <para>Nix now uses SQLite for its database.  This is faster and
    more flexible than the old <emphasis>ad hoc</emphasis> format.
    SQLite is also used to cache the manifests in
    <filename>/nix/var/nix/manifests</filename>, resulting in a
    significant speedup.</para>
  </listitem>

  <listitem>
    <para>Nix now has an search path for expressions.  The search path
    is set using the environment variable <envar>NIX_PATH</envar> and
    the <option>-I</option> command line option.  In Nix expressions,
    paths between angle brackets are used to specify files that must
    be looked up in the search path.  For instance, the expression
    <literal>&lt;nixpkgs/default.nix&gt;</literal> looks for a file
    <filename>nixpkgs/default.nix</filename> relative to every element
    in the search path.</para>
  </listitem>

  <listitem>
    <para>The new command <command>nix-build --run-env</command>
    builds all dependencies of a derivation, then starts a shell in an
    environment containing all variables from the derivation.  This is
    useful for reproducing the environment of a derivation for
    development.</para>
  </listitem>

  <listitem>
    <para>The new command <command>nix-store --verify-path</command>
    verifies that the contents of a store path have not
    changed.</para>
  </listitem>

  <listitem>
    <para>The new command <command>nix-store --print-env</command>
    prints out the environment of a derivation in a format that can be
    evaluated by a shell.</para>
  </listitem>

  <listitem>
    <para>Attribute names can now be arbitrary strings.  For instance,
    you can write <literal>{ "foo-1.2" = &#x2026;; "bla bla" = &#x2026;; }."bla
    bla"</literal>.</para>
  </listitem>

  <listitem>
    <para>Attribute selection can now provide a default value using
    the <literal>or</literal> operator.  For instance, the expression
    <literal>x.y.z or e</literal> evaluates to the attribute
    <literal>x.y.z</literal> if it exists, and <literal>e</literal>
    otherwise.</para>
  </listitem>

  <listitem>
    <para>The right-hand side of the <literal>?</literal> operator can
    now be an attribute path, e.g., <literal>attrs ?
    a.b.c</literal>.</para>
  </listitem>

  <listitem>
    <para>On Linux, Nix will now make files in the Nix store immutable
    on filesystems that support it.  This prevents accidental
    modification of files in the store by the root user.</para>
  </listitem>

  <listitem>
    <para>Nix has preliminary support for derivations with multiple
    outputs.  This is useful because it allows parts of a package to
    be deployed and garbage-collected separately.  For instance,
    development parts of a package such as header files or static
    libraries would typically not be part of the closure of an
    application, resulting in reduced disk usage and installation
    time.</para>
  </listitem>

  <listitem>
    <para>The Nix store garbage collector is faster and holds the
    global lock for a shorter amount of time.</para>
  </listitem>

  <listitem>
    <para>The option <option>--timeout</option> (corresponding to the
    configuration setting <literal>build-timeout</literal>) allows you
    to set an absolute timeout on builds &#x2014; if a build runs for more than
    the given number of seconds, it is terminated.  This is useful for
    recovering automatically from builds that are stuck in an infinite
    loop but keep producing output, and for which
    <literal>--max-silent-time</literal> is ineffective.</para>
  </listitem>

  <listitem>
    <para>Nix development has moved to GitHub (<link xlink:href="https://github.com/NixOS/nix"/>).</para>
  </listitem>

</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-0.16">

<title>Release 0.16 (2010-08-17)</title>

<para>This release has the following improvements:</para>

<itemizedlist>

  <listitem>
    <para>The Nix expression evaluator is now much faster in most
    cases: typically, <link xlink:href="http://www.mail-archive.com/nix-dev@cs.uu.nl/msg04113.html">3
    to 8 times compared to the old implementation</link>.  It also
    uses less memory.  It no longer depends on the ATerm
    library.</para>
  </listitem>

  <listitem>
    <para>
      Support for configurable parallelism inside builders.  Build
      scripts have always had the ability to perform multiple build
      actions in parallel (for instance, by running <command>make -j
      2</command>), but this was not desirable because the number of
      actions to be performed in parallel was not configurable.  Nix
      now has an option <option>--cores
      <replaceable>N</replaceable></option> as well as a configuration
      setting <varname>build-cores =
      <replaceable>N</replaceable></varname> that causes the
      environment variable <envar>NIX_BUILD_CORES</envar> to be set to
      <replaceable>N</replaceable> when the builder is invoked.  The
      builder can use this at its discretion to perform a parallel
      build, e.g., by calling <command>make -j
      <replaceable>N</replaceable></command>.  In Nixpkgs, this can be
      enabled on a per-package basis by setting the derivation
      attribute <varname>enableParallelBuilding</varname> to
      <literal>true</literal>.
    </para>
  </listitem>

  <listitem>
    <para><command>nix-store -q</command> now supports XML output
    through the <option>--xml</option> flag.</para>
  </listitem>

  <listitem>
    <para>Several bug fixes.</para>
  </listitem>

</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-0.15">

<title>Release 0.15 (2010-03-17)</title>

<para>This is a bug-fix release.  Among other things, it fixes
building on Mac OS X (Snow Leopard), and improves the contents of
<filename>/etc/passwd</filename> and <filename>/etc/group</filename>
in <literal>chroot</literal> builds.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-0.14">

<title>Release 0.14 (2010-02-04)</title>

<para>This release has the following improvements:</para>

<itemizedlist>

  <listitem>
    <para>The garbage collector now starts deleting garbage much
    faster than before.  It no longer determines liveness of all paths
    in the store, but does so on demand.</para>
  </listitem>

  <listitem>
    <para>Added a new operation, <command>nix-store --query
    --roots</command>, that shows the garbage collector roots that
    directly or indirectly point to the given store paths.</para>
  </listitem>

  <listitem>
    <para>Removed support for converting Berkeley DB-based Nix
    databases to the new schema.</para>
  </listitem>

  <listitem>
    <para>Removed the <option>--use-atime</option> and
    <option>--max-atime</option> garbage collector options.  They were
    not very useful in practice.</para>
  </listitem>

  <listitem>
    <para>On Windows, Nix now requires Cygwin 1.7.x.</para>
  </listitem>

  <listitem>
    <para>A few bug fixes.</para>
  </listitem>

</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-0.13">

<title>Release 0.13 (2009-11-05)</title>

<para>This is primarily a bug fix release.  It has some new
features:</para>

<itemizedlist>

  <listitem>
    <para>Syntactic sugar for writing nested attribute sets.  Instead of

<programlisting>
{
  foo = {
    bar = 123;
    xyzzy = true;
  };
  a = { b = { c = "d"; }; };
}
</programlisting>

    you can write

<programlisting>
{
  foo.bar = 123;
  foo.xyzzy = true;
  a.b.c = "d";
}
</programlisting>

    This is useful, for instance, in NixOS configuration files.</para>

  </listitem>

  <listitem>
    <para>Support for Nix channels generated by Hydra, the Nix-based
    continuous build system.  (Hydra generates NAR archives on the
    fly, so the size and hash of these archives isn&#x2019;t known in
    advance.)</para>
  </listitem>

  <listitem>
    <para>Support <literal>i686-linux</literal> builds directly on
    <literal>x86_64-linux</literal> Nix installations.  This is
    implemented using the <function>personality()</function> syscall,
    which causes <command>uname</command> to return
    <literal>i686</literal> in child processes.</para>
  </listitem>

  <listitem>
    <para>Various improvements to the <literal>chroot</literal>
    support.  Building in a <literal>chroot</literal> works quite well
    now.</para>
  </listitem>

  <listitem>
    <para>Nix no longer blocks if it tries to build a path and another
    process is already building the same path.  Instead it tries to
    build another buildable path first.  This improves
    parallelism.</para>
  </listitem>

  <listitem>
    <para>Support for large (&gt; 4 GiB) files in NAR archives.</para>
  </listitem>

  <listitem>
    <para>Various (performance) improvements to the remote build
    mechanism.</para>
  </listitem>

  <listitem>
    <para>New primops: <varname>builtins.addErrorContext</varname> (to
    add a string to stack traces &#x2014; useful for debugging),
    <varname>builtins.isBool</varname>,
    <varname>builtins.isString</varname>,
    <varname>builtins.isInt</varname>,
    <varname>builtins.intersectAttrs</varname>.</para>
  </listitem>

  <listitem>
    <para>OpenSolaris support (Sander van der Burg).</para>
  </listitem>

  <listitem>
    <para>Stack traces are no longer displayed unless the
    <option>--show-trace</option> option is used.</para>
  </listitem>

  <listitem>
    <para>The scoping rules for <literal>inherit
    (<replaceable>e</replaceable>) ...</literal> in recursive
    attribute sets have changed.  The expression
    <replaceable>e</replaceable> can now refer to the attributes
    defined in the containing set.</para>
  </listitem>

</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-0.12">

<title>Release 0.12 (2008-11-20)</title>

<itemizedlist>

  <listitem>
    <para>Nix no longer uses Berkeley DB to store Nix store metadata.
    The principal advantages of the new storage scheme are: it works
    properly over decent implementations of NFS (allowing Nix stores
    to be shared between multiple machines); no recovery is needed
    when a Nix process crashes; no write access is needed for
    read-only operations; no more running out of Berkeley DB locks on
    certain operations.</para>

    <para>You still need to compile Nix with Berkeley DB support if
    you want Nix to automatically convert your old Nix store to the
    new schema.  If you don&#x2019;t need this, you can build Nix with the
    <filename>configure</filename> option
    <option>--disable-old-db-compat</option>.</para>

    <para>After the automatic conversion to the new schema, you can
    delete the old Berkeley DB files:

    <screen>
$ cd /nix/var/nix/db
$ rm __db* log.* derivers references referrers reserved validpaths DB_CONFIG</screen>

    The new metadata is stored in the directories
    <filename>/nix/var/nix/db/info</filename> and
    <filename>/nix/var/nix/db/referrer</filename>.  Though the
    metadata is stored in human-readable plain-text files, they are
    not intended to be human-editable, as Nix is rather strict about
    the format.</para>

    <para>The new storage schema may or may not require less disk
    space than the Berkeley DB environment, mostly depending on the
    cluster size of your file system.  With 1 KiB clusters (which
    seems to be the <literal>ext3</literal> default nowadays) it
    usually takes up much less space.</para>
  </listitem>

  <listitem><para>There is a new substituter that copies paths
  directly from other (remote) Nix stores mounted somewhere in the
  filesystem.  For instance, you can speed up an installation by
  mounting some remote Nix store that already has the packages in
  question via NFS or <literal>sshfs</literal>.  The environment
  variable <envar>NIX_OTHER_STORES</envar> specifies the locations of
  the remote Nix directories,
  e.g. <literal>/mnt/remote-fs/nix</literal>.</para></listitem>

  <listitem><para>New <command>nix-store</command> operations
  <option>--dump-db</option> and <option>--load-db</option> to dump
  and reload the Nix database.</para></listitem>

  <listitem><para>The garbage collector has a number of new options to
  allow only some of the garbage to be deleted.  The option
  <option>--max-freed <replaceable>N</replaceable></option> tells the
  collector to stop after at least <replaceable>N</replaceable> bytes
  have been deleted.  The option <option>--max-links
  <replaceable>N</replaceable></option> tells it to stop after the
  link count on <filename>/nix/store</filename> has dropped below
  <replaceable>N</replaceable>.  This is useful for very large Nix
  stores on filesystems with a 32000 subdirectories limit (like
  <literal>ext3</literal>).  The option <option>--use-atime</option>
  causes store paths to be deleted in order of ascending last access
  time.  This allows non-recently used stuff to be deleted.  The
  option <option>--max-atime <replaceable>time</replaceable></option>
  specifies an upper limit to the last accessed time of paths that may
  be deleted.  For instance,

    <screen>
    $ nix-store --gc -v --max-atime $(date +%s -d "2 months ago")</screen>

  deletes everything that hasn&#x2019;t been accessed in two months.</para></listitem>

  <listitem><para><command>nix-env</command> now uses optimistic
  profile locking when performing an operation like installing or
  upgrading, instead of setting an exclusive lock on the profile.
  This allows multiple <command>nix-env -i / -u / -e</command>
  operations on the same profile in parallel.  If a
  <command>nix-env</command> operation sees at the end that the profile
  was changed in the meantime by another process, it will just
  restart.  This is generally cheap because the build results are
  still in the Nix store.</para></listitem>

  <listitem><para>The option <option>--dry-run</option> is now
  supported by <command>nix-store -r</command> and
  <command>nix-build</command>.</para></listitem>

  <listitem><para>The information previously shown by
  <option>--dry-run</option> (i.e., which derivations will be built
  and which paths will be substituted) is now always shown by
  <command>nix-env</command>, <command>nix-store -r</command> and
  <command>nix-build</command>.  The total download size of
  substitutable paths is now also shown.  For instance, a build will
  show something like

    <screen>
the following derivations will be built:
  /nix/store/129sbxnk5n466zg6r1qmq1xjv9zymyy7-activate-configuration.sh.drv
  /nix/store/7mzy971rdm8l566ch8hgxaf89x7lr7ik-upstart-jobs.drv
  ...
the following paths will be downloaded/copied (30.02 MiB):
  /nix/store/4m8pvgy2dcjgppf5b4cj5l6wyshjhalj-samba-3.2.4
  /nix/store/7h1kwcj29ip8vk26rhmx6bfjraxp0g4l-libunwind-0.98.6
  ...</screen>

  </para></listitem>

  <listitem><para>Language features:

    <itemizedlist>

      <listitem><para>@-patterns as in Haskell.  For instance, in a
      function definition

      <programlisting>f = args @ {x, y, z}: <replaceable>...</replaceable>;</programlisting>

      <varname>args</varname> refers to the argument as a whole, which
      is further pattern-matched against the attribute set pattern
      <literal>{x, y, z}</literal>.</para></listitem>

      <listitem><para>&#x201C;<literal>...</literal>&#x201D; (ellipsis) patterns.
      An attribute set pattern can now say <literal>...</literal>  at
      the end of the attribute name list to specify that the function
      takes <emphasis>at least</emphasis> the listed attributes, while
      ignoring additional attributes.  For instance,

      <programlisting>{stdenv, fetchurl, fuse, ...}: <replaceable>...</replaceable></programlisting>

      defines a function that accepts any attribute set that includes
      at least the three listed attributes.</para></listitem>

      <listitem><para>New primops:
      <varname>builtins.parseDrvName</varname> (split a package name
      string like <literal>"nix-0.12pre12876"</literal> into its name
      and version components, e.g. <literal>"nix"</literal> and
      <literal>"0.12pre12876"</literal>),
      <varname>builtins.compareVersions</varname> (compare two version
      strings using the same algorithm that <command>nix-env</command>
      uses), <varname>builtins.length</varname> (efficiently compute
      the length of a list), <varname>builtins.mul</varname> (integer
      multiplication), <varname>builtins.div</varname> (integer
      division).
      <!-- <varname>builtins.genericClosure</varname> -->
      </para></listitem>

    </itemizedlist>

  </para></listitem>

  <listitem><para><command>nix-prefetch-url</command> now supports
  <literal>mirror://</literal> URLs, provided that the environment
  variable <envar>NIXPKGS_ALL</envar> points at a Nixpkgs
  tree.</para></listitem>

  <listitem><para>Removed the commands
  <command>nix-pack-closure</command> and
  <command>nix-unpack-closure</command>.   You can do almost the same
  thing but much more efficiently by doing <literal>nix-store --export
  $(nix-store -qR <replaceable>paths</replaceable>) &gt; closure</literal> and
  <literal>nix-store --import &lt;
  closure</literal>.</para></listitem>

  <listitem><para>Lots of bug fixes, including a big performance bug in
  the handling of <literal>with</literal>-expressions.</para></listitem>

</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ssec-relnotes-0.11">

<title>Release 0.11 (2007-12-31)</title>

<para>Nix 0.11 has many improvements over the previous stable release.
The most important improvement is secure multi-user support.  It also
features many usability enhancements and language extensions, many of
them prompted by NixOS, the purely functional Linux distribution based
on Nix.  Here is an (incomplete) list:</para>


<itemizedlist>


  <listitem><para>Secure multi-user support.  A single Nix store can
  now be shared between multiple (possible untrusted) users.  This is
  an important feature for NixOS, where it allows non-root users to
  install software.  The old setuid method for sharing a store between
  multiple users has been removed.  Details for setting up a
  multi-user store can be found in the manual.</para></listitem>


  <listitem><para>The new command <command>nix-copy-closure</command>
  gives you an easy and efficient way to exchange software between
  machines.  It copies the missing parts of the closure of a set of
  store path to or from a remote machine via
  <command>ssh</command>.</para></listitem>


  <listitem><para>A new kind of string literal: strings between double
  single-quotes (<literal>''</literal>) have indentation
  &#x201C;intelligently&#x201D; removed.  This allows large strings (such as shell
  scripts or configuration file fragments in NixOS) to cleanly follow
  the indentation of the surrounding expression.  It also requires
  much less escaping, since <literal>''</literal> is less common in
  most languages than <literal>"</literal>.</para></listitem>


  <listitem><para><command>nix-env</command> <option>--set</option>
  modifies the current generation of a profile so that it contains
  exactly the specified derivation, and nothing else.  For example,
  <literal>nix-env -p /nix/var/nix/profiles/browser --set
  firefox</literal> lets the profile named
  <filename>browser</filename> contain just Firefox.</para></listitem>


  <listitem><para><command>nix-env</command> now maintains
  meta-information about installed packages in profiles.  The
  meta-information is the contents of the <varname>meta</varname>
  attribute of derivations, such as <varname>description</varname> or
  <varname>homepage</varname>.  The command <literal>nix-env -q --xml
  --meta</literal> shows all meta-information.</para></listitem>


  <listitem><para><command>nix-env</command> now uses the
  <varname>meta.priority</varname> attribute of derivations to resolve
  filename collisions between packages.  Lower priority values denote
  a higher priority.  For instance, the GCC wrapper package and the
  Binutils package in Nixpkgs both have a file
  <filename>bin/ld</filename>, so previously if you tried to install
  both you would get a collision.  Now, on the other hand, the GCC
  wrapper declares a higher priority than Binutils, so the former&#x2019;s
  <filename>bin/ld</filename> is symlinked in the user
  environment.</para></listitem>


  <listitem><para><command>nix-env -i / -u</command>: instead of
  breaking package ties by version, break them by priority and version
  number.  That is, if there are multiple packages with the same name,
  then pick the package with the highest priority, and only use the
  version if there are multiple packages with the same
  priority.</para>

  <para>This makes it possible to mark specific versions/variant in
  Nixpkgs more or less desirable than others.  A typical example would
  be a beta version of some package (e.g.,
  <literal>gcc-4.2.0rc1</literal>) which should not be installed even
  though it is the highest version, except when it is explicitly
  selected (e.g., <literal>nix-env -i
  gcc-4.2.0rc1</literal>).</para></listitem>


  <listitem><para><command>nix-env --set-flag</command> allows meta
  attributes of installed packages to be modified.  There are several
  attributes that can be usefully modified, because they affect the
  behaviour of <command>nix-env</command> or the user environment
  build script:

    <itemizedlist>

      <listitem><para><varname>meta.priority</varname> can be changed
      to resolve filename clashes (see above).</para></listitem>

      <listitem><para><varname>meta.keep</varname> can be set to
      <literal>true</literal> to prevent the package from being
      upgraded or replaced.  Useful if you want to hang on to an older
      version of a package.</para></listitem>

      <listitem><para><varname>meta.active</varname> can be set to
      <literal>false</literal> to &#x201C;disable&#x201D; the package.  That is, no
      symlinks will be generated to the files of the package, but it
      remains part of the profile (so it won&#x2019;t be garbage-collected).
      Set it back to <literal>true</literal> to re-enable the
      package.</para></listitem>

    </itemizedlist>

  </para></listitem>


  <listitem><para><command>nix-env -q</command> now has a flag
  <option>--prebuilt-only</option> (<option>-b</option>) that causes
  <command>nix-env</command> to show only those derivations whose
  output is already in the Nix store or that can be substituted (i.e.,
  downloaded from somewhere).  In other words, it shows the packages
  that can be installed &#x201C;quickly&#x201D;, i.e., don&#x2019;t need to be built from
  source.  The <option>-b</option> flag is also available in
  <command>nix-env -i</command> and <command>nix-env -u</command> to
  filter out derivations for which no pre-built binary is
  available.</para></listitem>


  <listitem><para>The new option <option>--argstr</option> (in
  <command>nix-env</command>, <command>nix-instantiate</command> and
  <command>nix-build</command>) is like <option>--arg</option>, except
  that the value is a string.  For example, <literal>--argstr system
  i686-linux</literal> is equivalent to <literal>--arg system
  \"i686-linux\"</literal> (note that <option>--argstr</option>
  prevents annoying quoting around shell arguments).</para></listitem>


  <listitem><para><command>nix-store</command> has a new operation
  <option>--read-log</option> (<option>-l</option>)
  <parameter>paths</parameter> that shows the build log of the given
  paths.</para></listitem>


  <!--
  <listitem><para>TODO: semantic cleanups of string concatenation
  etc. (mostly in r6740).</para></listitem>
  -->


  <listitem><para>Nix now uses Berkeley DB 4.5.  The database is
  upgraded automatically, but you should be careful not to use old
  versions of Nix that still use Berkeley DB 4.4.</para></listitem>


  <!-- foo
  <listitem><para>TODO: option <option>- -reregister</option> in
  <command>nix-store - -register-validity</command>.</para></listitem>
  -->


  <listitem><para>The option <option>--max-silent-time</option>
  (corresponding to the configuration setting
  <literal>build-max-silent-time</literal>) allows you to set a
  timeout on builds &#x2014; if a build produces no output on
  <literal>stdout</literal> or <literal>stderr</literal> for the given
  number of seconds, it is terminated.  This is useful for recovering
  automatically from builds that are stuck in an infinite
  loop.</para></listitem>


  <listitem><para><command>nix-channel</command>: each subscribed
  channel is its own attribute in the top-level expression generated
  for the channel.  This allows disambiguation (e.g. <literal>nix-env
  -i -A nixpkgs_unstable.firefox</literal>).</para></listitem>


  <listitem><para>The substitutes table has been removed from the
  database.  This makes operations such as <command>nix-pull</command>
  and <command>nix-channel --update</command> much, much
  faster.</para></listitem>


  <listitem><para><command>nix-pull</command> now supports
  bzip2-compressed manifests.  This speeds up
  channels.</para></listitem>


  <listitem><para><command>nix-prefetch-url</command> now has a
  limited form of caching.  This is used by
  <command>nix-channel</command> to prevent unnecessary downloads when
  the channel hasn&#x2019;t changed.</para></listitem>


  <listitem><para><command>nix-prefetch-url</command> now by default
  computes the SHA-256 hash of the file instead of the MD5 hash.  In
  calls to <function>fetchurl</function> you should pass the
  <literal>sha256</literal> attribute instead of
  <literal>md5</literal>.  You can pass either a hexadecimal or a
  base-32 encoding of the hash.</para></listitem>


  <listitem><para>Nix can now perform builds in an automatically
  generated &#x201C;chroot&#x201D;.  This prevents a builder from accessing stuff
  outside of the Nix store, and thus helps ensure purity.  This is an
  experimental feature.</para></listitem>


  <listitem><para>The new command <command>nix-store
  --optimise</command> reduces Nix store disk space usage by finding
  identical files in the store and hard-linking them to each other.
  It typically reduces the size of the store by something like
  25-35%.</para></listitem>


  <listitem><para><filename>~/.nix-defexpr</filename> can now be a
  directory, in which case the Nix expressions in that directory are
  combined into an attribute set, with the file names used as the
  names of the attributes.  The command <command>nix-env
  --import</command> (which set the
  <filename>~/.nix-defexpr</filename> symlink) is
  removed.</para></listitem>


  <listitem><para>Derivations can specify the new special attribute
  <varname>allowedReferences</varname> to enforce that the references
  in the output of a derivation are a subset of a declared set of
  paths.  For example, if <varname>allowedReferences</varname> is an
  empty list, then the output must not have any references.  This is
  used in NixOS to check that generated files such as initial ramdisks
  for booting Linux don&#x2019;t have any dependencies.</para></listitem>


  <listitem><para>The new attribute
  <varname>exportReferencesGraph</varname> allows builders access to
  the references graph of their inputs.  This is used in NixOS for
  tasks such as generating ISO-9660 images that contain a Nix store
  populated with the closure of certain paths.</para></listitem>


  <listitem><para>Fixed-output derivations (like
  <function>fetchurl</function>) can define the attribute
  <varname>impureEnvVars</varname> to allow external environment
  variables to be passed to builders.  This is used in Nixpkgs to
  support proxy configuration, among other things.</para></listitem>


  <listitem><para>Several new built-in functions:
  <function>builtins.attrNames</function>,
  <function>builtins.filterSource</function>,
  <function>builtins.isAttrs</function>,
  <function>builtins.isFunction</function>,
  <function>builtins.listToAttrs</function>,
  <function>builtins.stringLength</function>,
  <function>builtins.sub</function>,
  <function>builtins.substring</function>,
  <function>throw</function>,
  <function>builtins.trace</function>,
  <function>builtins.readFile</function>.</para></listitem>


</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-relnotes-0.10.1">

<title>Release 0.10.1 (2006-10-11)</title>

<para>This release fixes two somewhat obscure bugs that occur when
evaluating Nix expressions that are stored inside the Nix store
(<literal>NIX-67</literal>).  These do not affect most users.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-relnotes-0.10">

<title>Release 0.10 (2006-10-06)</title>

<note><para>This version of Nix uses Berkeley DB 4.4 instead of 4.3.
The database is upgraded automatically, but you should be careful not
to use old versions of Nix that still use Berkeley DB 4.3.  In
particular, if you use a Nix installed through Nix, you should run

<screen>
$ nix-store --clear-substitutes</screen>

first.</para></note>

<warning><para>Also, the database schema has changed slighted to fix a
performance issue (see below).  When you run any Nix 0.10 command for
the first time, the database will be upgraded automatically.  This is
irreversible.</para></warning>

<itemizedlist>


  <!-- Usability / features -->


  <listitem><para><command>nix-env</command> usability improvements:

    <itemizedlist>

      <listitem><para>An option <option>--compare-versions</option>
      (or <option>-c</option>) has been added to <command>nix-env
      --query</command> to allow you to compare installed versions of
      packages to available versions, or vice versa.  An easy way to
      see if you are up to date with what&#x2019;s in your subscribed
      channels is <literal>nix-env -qc \*</literal>.</para></listitem>

      <listitem><para><literal>nix-env --query</literal> now takes as
      arguments a list of package names about which to show
      information, just like <option>--install</option>, etc.: for
      example, <literal>nix-env -q gcc</literal>.  Note that to show
      all derivations, you need to specify
      <literal>\*</literal>.</para></listitem>

      <listitem><para><literal>nix-env -i
      <replaceable>pkgname</replaceable></literal> will now install
      the highest available version of
      <replaceable>pkgname</replaceable>, rather than installing all
      available versions (which would probably give collisions)
      (<literal>NIX-31</literal>).</para></listitem>

      <listitem><para><literal>nix-env (-i|-u) --dry-run</literal> now
      shows exactly which missing paths will be built or
      substituted.</para></listitem>

      <listitem><para><literal>nix-env -qa --description</literal>
      shows human-readable descriptions of packages, provided that
      they have a <literal>meta.description</literal> attribute (which
      most packages in Nixpkgs don&#x2019;t have yet).</para></listitem>

    </itemizedlist>

  </para></listitem>


  <listitem><para>New language features:

    <itemizedlist>

      <listitem><para>Reference scanning (which happens after each
      build) is much faster and takes a constant amount of
      memory.</para></listitem>

      <listitem><para>String interpolation.  Expressions like

<programlisting>
"--with-freetype2-library=" + freetype + "/lib"</programlisting>

      can now be written as

<programlisting>
"--with-freetype2-library=${freetype}/lib"</programlisting>

      You can write arbitrary expressions within
      <literal>${<replaceable>...</replaceable>}</literal>, not just
      identifiers.</para></listitem>

      <listitem><para>Multi-line string literals.</para></listitem>

      <listitem><para>String concatenations can now involve
      derivations, as in the example <code>"--with-freetype2-library="
      + freetype + "/lib"</code>.  This was not previously possible
      because we need to register that a derivation that uses such a
      string is dependent on <literal>freetype</literal>.  The
      evaluator now properly propagates this information.
      Consequently, the subpath operator (<literal>~</literal>) has
      been deprecated.</para></listitem>

      <listitem><para>Default values of function arguments can now
      refer to other function arguments; that is, all arguments are in
      scope in the default values
      (<literal>NIX-45</literal>).</para></listitem>

      <!--
      <listitem><para>TODO: domain checks (r5895).</para></listitem>
      -->

      <listitem><para>Lots of new built-in primitives, such as
      functions for list manipulation and integer arithmetic.  See the
      manual for a complete list.  All primops are now available in
      the set <varname>builtins</varname>, allowing one to test for
      the availability of primop in a backwards-compatible
      way.</para></listitem>

      <listitem><para>Real let-expressions: <literal>let x = ...;
      ... z = ...; in ...</literal>.</para></listitem>

    </itemizedlist>

  </para></listitem>


  <listitem><para>New commands <command>nix-pack-closure</command> and
  <command>nix-unpack-closure</command> than can be used to easily
  transfer a store path with all its dependencies to another machine.
  Very convenient whenever you have some package on your machine and
  you want to copy it somewhere else.</para></listitem>


  <listitem><para>XML support:

    <itemizedlist>

      <listitem><para><literal>nix-env -q --xml</literal> prints the
      installed or available packages in an XML representation for
      easy processing by other tools.</para></listitem>

      <listitem><para><literal>nix-instantiate --eval-only
      --xml</literal> prints an XML representation of the resulting
      term.  (The new flag <option>--strict</option> forces &#x2018;deep&#x2019;
      evaluation of the result, i.e., list elements and attributes are
      evaluated recursively.)</para></listitem>

      <listitem><para>In Nix expressions, the primop
      <function>builtins.toXML</function> converts a term to an XML
      representation.  This is primarily useful for passing structured
      information to builders.</para></listitem>

    </itemizedlist>

  </para></listitem>


  <listitem><para>You can now unambiguously specify which derivation to
  build or install in <command>nix-env</command>,
  <command>nix-instantiate</command> and <command>nix-build</command>
  using the <option>--attr</option> / <option>-A</option> flags, which
  takes an attribute name as argument.  (Unlike symbolic package names
  such as <literal>subversion-1.4.0</literal>, attribute names in an
  attribute set are unique.)  For instance, a quick way to perform a
  test build of a package in Nixpkgs is <literal>nix-build
  pkgs/top-level/all-packages.nix -A
  <replaceable>foo</replaceable></literal>.  <literal>nix-env -q
  --attr</literal> shows the attribute names corresponding to each
  derivation.</para></listitem>


  <listitem><para>If the top-level Nix expression used by
  <command>nix-env</command>, <command>nix-instantiate</command> or
  <command>nix-build</command> evaluates to a function whose arguments
  all have default values, the function will be called automatically.
  Also, the new command-line switch <option>--arg
  <replaceable>name</replaceable>
  <replaceable>value</replaceable></option> can be used to specify
  function arguments on the command line.</para></listitem>


  <listitem><para><literal>nix-install-package --url
  <replaceable>URL</replaceable></literal> allows a package to be
  installed directly from the given URL.</para></listitem>


  <listitem><para>Nix now works behind an HTTP proxy server; just set
  the standard environment variables <envar>http_proxy</envar>,
  <envar>https_proxy</envar>, <envar>ftp_proxy</envar> or
  <envar>all_proxy</envar> appropriately.  Functions such as
  <function>fetchurl</function> in Nixpkgs also respect these
  variables.</para></listitem>


  <listitem><para><literal>nix-build -o
  <replaceable>symlink</replaceable></literal> allows the symlink to
  the build result to be named something other than
  <literal>result</literal>.</para></listitem>


  <!-- Stability / performance / etc. -->


  <listitem><para>Platform support:

    <itemizedlist>

      <listitem><para>Support for 64-bit platforms, provided a <link xlink:href="http://bugzilla.sen.cwi.nl:8080/show_bug.cgi?id=606">suitably
      patched ATerm library</link> is used.  Also, files larger than 2
      GiB are now supported.</para></listitem>

      <listitem><para>Added support for Cygwin (Windows,
      <literal>i686-cygwin</literal>), Mac OS X on Intel
      (<literal>i686-darwin</literal>) and Linux on PowerPC
      (<literal>powerpc-linux</literal>).</para></listitem>

      <listitem><para>Users of SMP and multicore machines will
      appreciate that the number of builds to be performed in parallel
      can now be specified in the configuration file in the
      <literal>build-max-jobs</literal> setting.</para></listitem>

    </itemizedlist>

  </para></listitem>


  <listitem><para>Garbage collector improvements:

    <itemizedlist>

      <listitem><para>Open files (such as running programs) are now
      used as roots of the garbage collector.  This prevents programs
      that have been uninstalled from being garbage collected while
      they are still running.  The script that detects these
      additional runtime roots
      (<filename>find-runtime-roots.pl</filename>) is inherently
      system-specific, but it should work on Linux and on all
      platforms that have the <command>lsof</command>
      utility.</para></listitem>

      <listitem><para><literal>nix-store --gc</literal>
      (a.k.a. <command>nix-collect-garbage</command>) prints out the
      number of bytes freed on standard output.  <literal>nix-store
      --gc --print-dead</literal> shows how many bytes would be freed
      by an actual garbage collection.</para></listitem>

      <listitem><para><literal>nix-collect-garbage -d</literal>
      removes all old generations of <emphasis>all</emphasis> profiles
      before calling the actual garbage collector (<literal>nix-store
      --gc</literal>).  This is an easy way to get rid of all old
      packages in the Nix store.</para></listitem>

      <listitem><para><command>nix-store</command> now has an
      operation <option>--delete</option> to delete specific paths
      from the Nix store.  It won&#x2019;t delete reachable (non-garbage)
      paths unless <option>--ignore-liveness</option> is
      specified.</para></listitem>

    </itemizedlist>

  </para></listitem>


  <listitem><para>Berkeley DB 4.4&#x2019;s process registry feature is used
  to recover from crashed Nix processes.</para></listitem>

  <!--  <listitem><para>TODO: shared stores.</para></listitem> -->

  <listitem><para>A performance issue has been fixed with the
  <literal>referer</literal> table, which stores the inverse of the
  <literal>references</literal> table (i.e., it tells you what store
  paths refer to a given path).  Maintaining this table could take a
  quadratic amount of time, as well as a quadratic amount of Berkeley
  DB log file space (in particular when running the garbage collector)
  (<literal>NIX-23</literal>).</para></listitem>

  <listitem><para>Nix now catches the <literal>TERM</literal> and
  <literal>HUP</literal> signals in addition to the
  <literal>INT</literal> signal.  So you can now do a <literal>killall
  nix-store</literal> without triggering a database
  recovery.</para></listitem>

  <listitem><para><command>bsdiff</command> updated to version
  4.3.</para></listitem>

  <listitem><para>Substantial performance improvements in expression
  evaluation and <literal>nix-env -qa</literal>, all thanks to <link xlink:href="http://valgrind.org/">Valgrind</link>.  Memory use has
  been reduced by a factor 8 or so.  Big speedup by memoisation of
  path hashing.</para></listitem>

  <listitem><para>Lots of bug fixes, notably:

    <itemizedlist>

      <listitem><para>Make sure that the garbage collector can run
      successfully when the disk is full
      (<literal>NIX-18</literal>).</para></listitem>

      <listitem><para><command>nix-env</command> now locks the profile
      to prevent races between concurrent <command>nix-env</command>
      operations on the same profile
      (<literal>NIX-7</literal>).</para></listitem>

      <listitem><para>Removed misleading messages from
      <literal>nix-env -i</literal> (e.g., <literal>installing
      `foo'</literal> followed by <literal>uninstalling
      `foo'</literal>) (<literal>NIX-17</literal>).</para></listitem>

    </itemizedlist>

  </para></listitem>

  <listitem><para>Nix source distributions are a lot smaller now since
  we no longer include a full copy of the Berkeley DB source
  distribution (but only the bits we need).</para></listitem>

  <listitem><para>Header files are now installed so that external
  programs can use the Nix libraries.</para></listitem>

</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-relnotes-0.9.2">

<title>Release 0.9.2 (2005-09-21)</title>

<para>This bug fix release fixes two problems on Mac OS X:

<itemizedlist>

  <listitem><para>If Nix was linked against statically linked versions
  of the ATerm or Berkeley DB library, there would be dynamic link
  errors at runtime.</para></listitem>

  <listitem><para><command>nix-pull</command> and
  <command>nix-push</command> intermittently failed due to race
  conditions involving pipes and child processes with error messages
  such as <literal>open2: open(GLOB(0x180b2e4), &gt;&amp;=9) failed: Bad
  file descriptor at /nix/bin/nix-pull line 77</literal> (issue
  <literal>NIX-14</literal>).</para></listitem>

</itemizedlist>

</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-relnotes-0.9.1">

<title>Release 0.9.1 (2005-09-20)</title>

<para>This bug fix release addresses a problem with the ATerm library
when the <option>--with-aterm</option> flag in
<command>configure</command> was <emphasis>not</emphasis> used.</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-relnotes-0.9">

<title>Release 0.9 (2005-09-16)</title>

<para>NOTE: this version of Nix uses Berkeley DB 4.3 instead of 4.2.
The database is upgraded automatically, but you should be careful not
to use old versions of Nix that still use Berkeley DB 4.2.  In
particular, if you use a Nix installed through Nix, you should run

<screen>
$ nix-store --clear-substitutes</screen>

first.</para>


<itemizedlist>

  <listitem><para>Unpacking of patch sequences is much faster now
  since we no longer do redundant unpacking and repacking of
  intermediate paths.</para></listitem>

  <listitem><para>Nix now uses Berkeley DB 4.3.</para></listitem>

  <listitem><para>The <function>derivation</function> primitive is
  lazier.  Attributes of dependent derivations can mutually refer to
  each other (as long as there are no data dependencies on the
  <varname>outPath</varname> and <varname>drvPath</varname> attributes
  computed by <function>derivation</function>).</para>

  <para>For example, the expression <literal>derivation
  attrs</literal> now evaluates to (essentially)

  <programlisting>
attrs // {
  type = "derivation";
  outPath = derivation! attrs;
  drvPath = derivation! attrs;
}</programlisting>

  where <function>derivation!</function> is a primop that does the
  actual derivation instantiation (i.e., it does what
  <function>derivation</function> used to do).  The advantage is that
  it allows commands such as <command>nix-env -qa</command> and
  <command>nix-env -i</command> to be much faster since they no longer
  need to instantiate all derivations, just the
  <varname>name</varname> attribute.</para>

  <para>Also, it allows derivations to cyclically reference each
  other, for example,

  <programlisting>
webServer = derivation {
  ...
  hostName = "svn.cs.uu.nl";
  services = [svnService];
};
 
svnService = derivation {
  ...
  hostName = webServer.hostName;
};</programlisting>

  Previously, this would yield a black hole (infinite recursion).</para>

  </listitem>

  <listitem><para><command>nix-build</command> now defaults to using
  <filename>./default.nix</filename> if no Nix expression is
  specified.</para></listitem>

  <listitem><para><command>nix-instantiate</command>, when applied to
  a Nix expression that evaluates to a function, will call the
  function automatically if all its arguments have
  defaults.</para></listitem>

  <listitem><para>Nix now uses libtool to build dynamic libraries.
  This reduces the size of executables.</para></listitem>

  <listitem><para>A new list concatenation operator
  <literal>++</literal>.  For example, <literal>[1 2 3] ++ [4 5
  6]</literal> evaluates to <literal>[1 2 3 4 5
  6]</literal>.</para></listitem>

  <listitem><para>Some currently undocumented primops to support
  low-level build management using Nix (i.e., using Nix as a Make
  replacement).  See the commit messages for <literal>r3578</literal>
  and <literal>r3580</literal>.</para></listitem>

  <listitem><para>Various bug fixes and performance
  improvements.</para></listitem>

</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-relnotes-0.8.1">

<title>Release 0.8.1 (2005-04-13)</title>

<para>This is a bug fix release.</para>

<itemizedlist>

  <listitem><para>Patch downloading was broken.</para></listitem>

  <listitem><para>The garbage collector would not delete paths that
  had references from invalid (but substitutable)
  paths.</para></listitem>

</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-relnotes-0.8">

<title>Release 0.8 (2005-04-11)</title>

<para>NOTE: the hashing scheme in Nix 0.8 changed (as detailed below).
As a result, <command>nix-pull</command> manifests and channels built
for Nix 0.7 and below will not work anymore.  However, the Nix
expression language has not changed, so you can still build from
source.  Also, existing user environments continue to work.  Nix 0.8
will automatically upgrade the database schema of previous
installations when it is first run.</para>

<para>If you get the error message

<screen>
you have an old-style manifest `/nix/var/nix/manifests/[...]'; please
delete it</screen>

you should delete previously downloaded manifests:

<screen>
$ rm /nix/var/nix/manifests/*</screen>

If <command>nix-channel</command> gives the error message

<screen>
manifest `http://catamaran.labs.cs.uu.nl/dist/nix/channels/[channel]/MANIFEST'
is too old (i.e., for Nix &lt;= 0.7)</screen>

then you should unsubscribe from the offending channel
(<command>nix-channel --remove
<replaceable>URL</replaceable></command>; leave out
<literal>/MANIFEST</literal>), and subscribe to the same URL, with
<literal>channels</literal> replaced by <literal>channels-v3</literal>
(e.g., <link xlink:href="http://catamaran.labs.cs.uu.nl/dist/nix/channels-v3/nixpkgs-unstable"/>).</para>

<para>Nix 0.8 has the following improvements:

<itemizedlist>

  <listitem><para>The cryptographic hashes used in store paths are now
  160 bits long, but encoded in base-32 so that they are still only 32
  characters long (e.g.,
  <filename>/nix/store/csw87wag8bqlqk7ipllbwypb14xainap-atk-1.9.0</filename>).
  (This is actually a 160 bit truncation of a SHA-256
  hash.)</para></listitem>

  <listitem><para>Big cleanups and simplifications of the basic store
  semantics.  The notion of &#x201C;closure store expressions&#x201D; is gone (and
  so is the notion of &#x201C;successors&#x201D;); the file system references of a
  store path are now just stored in the database.</para>

  <para>For instance, given any store path, you can query its closure:

  <screen>
$ nix-store -qR $(which firefox)
... lots of paths ...</screen>

  Also, Nix now remembers for each store path the derivation that
  built it (the &#x201C;deriver&#x201D;):

  <screen>
$ nix-store -qR $(which firefox)
/nix/store/4b0jx7vq80l9aqcnkszxhymsf1ffa5jd-firefox-1.0.1.drv</screen>

  So to see the build-time dependencies, you can do

  <screen>
$ nix-store -qR $(nix-store -qd $(which firefox))</screen>

  or, in a nicer format:

  <screen>
$ nix-store -q --tree $(nix-store -qd $(which firefox))</screen>

  </para>

  <para>File system references are also stored in reverse.  For
  instance, you can query all paths that directly or indirectly use a
  certain Glibc:

  <screen>
$ nix-store -q --referrers-closure \
    /nix/store/8lz9yc6zgmc0vlqmn2ipcpkjlmbi51vv-glibc-2.3.4</screen>

  </para>

  </listitem>

  <listitem><para>The concept of fixed-output derivations has been
  formalised.  Previously, functions such as
  <function>fetchurl</function> in Nixpkgs used a hack (namely,
  explicitly specifying a store path hash) to prevent changes to, say,
  the URL of the file from propagating upwards through the dependency
  graph, causing rebuilds of everything.  This can now be done cleanly
  by specifying the <varname>outputHash</varname> and
  <varname>outputHashAlgo</varname> attributes.  Nix itself checks
  that the content of the output has the specified hash.  (This is
  important for maintaining certain invariants necessary for future
  work on secure shared stores.)</para></listitem>

  <listitem><para>One-click installation :-) It is now possible to
  install any top-level component in Nixpkgs directly, through the web
  &#x2014; see, e.g., <link xlink:href="http://catamaran.labs.cs.uu.nl/dist/nixpkgs-0.8/"/>.
  All you have to do is associate
  <filename>/nix/bin/nix-install-package</filename> with the MIME type
  <literal>application/nix-package</literal> (or the extension
  <filename>.nixpkg</filename>), and clicking on a package link will
  cause it to be installed, with all appropriate dependencies.  If you
  just want to install some specific application, this is easier than
  subscribing to a channel.</para></listitem>

  <listitem><para><command>nix-store -r
  <replaceable>PATHS</replaceable></command> now builds all the
  derivations PATHS in parallel.  Previously it did them sequentially
  (though exploiting possible parallelism between subderivations).
  This is nice for build farms.</para></listitem>

  <listitem><para><command>nix-channel</command> has new operations
  <option>--list</option> and
  <option>--remove</option>.</para></listitem>

  <listitem><para>New ways of installing components into user
  environments:

  <itemizedlist>

    <listitem><para>Copy from another user environment:

    <screen>
$ nix-env -i --from-profile .../other-profile firefox</screen>

    </para></listitem>

    <listitem><para>Install a store derivation directly (bypassing the
    Nix expression language entirely):

    <screen>
$ nix-env -i /nix/store/z58v41v21xd3...-aterm-2.3.1.drv</screen>

    (This is used to implement <command>nix-install-package</command>,
    which is therefore immune to evolution in the Nix expression
    language.)</para></listitem>

    <listitem><para>Install an already built store path directly:

    <screen>
$ nix-env -i /nix/store/hsyj5pbn0d9i...-aterm-2.3.1</screen>

    </para></listitem>

    <listitem><para>Install the result of a Nix expression specified
    as a command-line argument:

    <screen>
$ nix-env -f .../i686-linux.nix -i -E 'x: x.firefoxWrapper'</screen>

    The difference with the normal installation mode is that
    <option>-E</option> does not use the <varname>name</varname>
    attributes of derivations.  Therefore, this can be used to
    disambiguate multiple derivations with the same
    name.</para></listitem>

  </itemizedlist></para></listitem>

  <listitem><para>A hash of the contents of a store path is now stored
  in the database after a successful build.  This allows you to check
  whether store paths have been tampered with: <command>nix-store
  --verify --check-contents</command>.</para></listitem>

  <listitem>

    <para>Implemented a concurrent garbage collector.  It is now
    always safe to run the garbage collector, even if other Nix
    operations are happening simultaneously.</para>

    <para>However, there can still be GC races if you use
    <command>nix-instantiate</command> and <command>nix-store
    --realise</command> directly to build things.  To prevent races,
    use the <option>--add-root</option> flag of those commands.</para>

  </listitem>

  <listitem><para>The garbage collector now finally deletes paths in
  the right order (i.e., topologically sorted under the &#x201C;references&#x201D;
  relation), thus making it safe to interrupt the collector without
  risking a store that violates the closure
  invariant.</para></listitem>

  <listitem><para>Likewise, the substitute mechanism now downloads
  files in the right order, thus preserving the closure invariant at
  all times.</para></listitem>

  <listitem><para>The result of <command>nix-build</command> is now
  registered as a root of the garbage collector.  If the
  <filename>./result</filename> link is deleted, the GC root
  disappears automatically.</para></listitem>

  <listitem>

    <para>The behaviour of the garbage collector can be changed
    globally by setting options in
    <filename>/nix/etc/nix/nix.conf</filename>.

    <itemizedlist>

      <listitem><para><literal>gc-keep-derivations</literal> specifies
      whether deriver links should be followed when searching for live
      paths.</para></listitem>

      <listitem><para><literal>gc-keep-outputs</literal> specifies
      whether outputs of derivations should be followed when searching
      for live paths.</para></listitem>

      <listitem><para><literal>env-keep-derivations</literal>
      specifies whether user environments should store the paths of
      derivations when they are added (thus keeping the derivations
      alive).</para></listitem>

    </itemizedlist>

  </para></listitem>

  <listitem><para>New <command>nix-env</command> query flags
  <option>--drv-path</option> and
  <option>--out-path</option>.</para></listitem>

  <listitem><para><command>fetchurl</command> allows SHA-1 and SHA-256
  in addition to MD5.  Just specify the attribute
  <varname>sha1</varname> or <varname>sha256</varname> instead of
  <varname>md5</varname>.</para></listitem>

  <listitem><para>Manual updates.</para></listitem>

</itemizedlist>

</para>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-relnotes-0.7">

<title>Release 0.7 (2005-01-12)</title>

<itemizedlist>

  <listitem><para>Binary patching.  When upgrading components using
  pre-built binaries (through nix-pull / nix-channel), Nix can
  automatically download and apply binary patches to already installed
  components instead of full downloads.  Patching is &#x201C;smart&#x201D;: if there
  is a <emphasis>sequence</emphasis> of patches to an installed
  component, Nix will use it.  Patches are currently generated
  automatically between Nixpkgs (pre-)releases.</para></listitem>

  <listitem><para>Simplifications to the substitute
  mechanism.</para></listitem>

  <listitem><para>Nix-pull now stores downloaded manifests in
  <filename>/nix/var/nix/manifests</filename>.</para></listitem>

  <listitem><para>Metadata on files in the Nix store is canonicalised
  after builds: the last-modified timestamp is set to 0 (00:00:00
  1/1/1970), the mode is set to 0444 or 0555 (readable and possibly
  executable by all; setuid/setgid bits are dropped), and the group is
  set to the default.  This ensures that the result of a build and an
  installation through a substitute is the same; and that timestamp
  dependencies are revealed.</para></listitem>

</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-relnotes-0.6">

<title>Release 0.6 (2004-11-14)</title>

<itemizedlist>

  <listitem>
    <para>Rewrite of the normalisation engine.

    <itemizedlist>

      <listitem><para>Multiple builds can now be performed in parallel
      (option <option>-j</option>).</para></listitem>

      <listitem><para>Distributed builds.  Nix can now call a shell
      script to forward builds to Nix installations on remote
      machines, which may or may not be of the same platform
      type.</para></listitem>

      <listitem><para>Option <option>--fallback</option> allows
      recovery from broken substitutes.</para></listitem>

      <listitem><para>Option <option>--keep-going</option> causes
      building of other (unaffected) derivations to continue if one
      failed.</para></listitem>

    </itemizedlist>

    </para>

  </listitem>

  <listitem><para>Improvements to the garbage collector (i.e., it
  should actually work now).</para></listitem>

  <listitem><para>Setuid Nix installations allow a Nix store to be
  shared among multiple users.</para></listitem>

  <listitem><para>Substitute registration is much faster
  now.</para></listitem>

  <listitem><para>A utility <command>nix-build</command> to build a
  Nix expression and create a symlink to the result int the current
  directory; useful for testing Nix derivations.</para></listitem>

  <listitem><para>Manual updates.</para></listitem>

  <listitem>

    <para><command>nix-env</command> changes:

    <itemizedlist>

      <listitem><para>Derivations for other platforms are filtered out
      (which can be overridden using
      <option>--system-filter</option>).</para></listitem>

      <listitem><para><option>--install</option> by default now
      uninstall previous derivations with the same
      name.</para></listitem>

      <listitem><para><option>--upgrade</option> allows upgrading to a
      specific version.</para></listitem>

      <listitem><para>New operation
      <option>--delete-generations</option> to remove profile
      generations (necessary for effective garbage
      collection).</para></listitem>

      <listitem><para>Nicer output (sorted,
      columnised).</para></listitem>

    </itemizedlist>

    </para>

  </listitem>

  <listitem><para>More sensible verbosity levels all around (builder
  output is now shown always, unless <option>-Q</option> is
  given).</para></listitem>

  <listitem>

    <para>Nix expression language changes:

    <itemizedlist>

      <listitem><para>New language construct: <literal>with
      <replaceable>E1</replaceable>;
      <replaceable>E2</replaceable></literal> brings all attributes
      defined in the attribute set <replaceable>E1</replaceable> in
      scope in <replaceable>E2</replaceable>.</para></listitem>

      <listitem><para>Added a <function>map</function>
      function.</para></listitem>

      <listitem><para>Various new operators (e.g., string
      concatenation).</para></listitem>

    </itemizedlist>

    </para>

  </listitem>

  <listitem><para>Expression evaluation is much
  faster.</para></listitem>

  <listitem><para>An Emacs mode for editing Nix expressions (with
  syntax highlighting and indentation) has been
  added.</para></listitem>

  <listitem><para>Many bug fixes.</para></listitem>

</itemizedlist>

</section>
<section xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xi="http://www.w3.org/2001/XInclude" version="5.0" xml:id="ch-relnotes-0.5">

<title>Release 0.5 and earlier</title>

<para>Please refer to the Subversion commit log messages.</para>

</section>

</appendix>

<!--
<appendix>
    <title>Nix Release Notes</title>
    <xi:include href="release-notes/release-notes.xml"
                xpointer="xmlns(x=http://docbook.org/ns/docbook)xpointer(x:article/x:section)" />
  </appendix>
-->

</book>