This is a major closure size reduction on Darwin, and probably a less
significant one on Linux. On darwin, retaining the compiler means adding
clang and its dependency llvm to the perl closure, which gives us ~400MB
of extra stuff. Considering that Nix itself depends on this version of
perl, that makes cutting a new Nix release rather unpleasaont Darwin.
After this patch, I was able to get the `nixUnstable` closure down to
21MB after feeding it into a .tar.xz (123MB before compression). There's
still room for improvement but this should carry us over until we split
outputs.
Add Twisted as build input so that we can continue to have Python
support. (./configure disables Python support unless it finds the
'trial' program, from Twisted.) I don't know whether upstream intended
that, because it seems perfectly fine to run thrift + Python without
Twisted. (Only the TTwisted transport uses Twisted...)
Ah, Thrift use Twisted in its unit tests. Even when we pass
--enable-tests=no to ./configure :-D
This update was generated by hackage2nix v20151217-5-ged07a8e using the following inputs:
- Nixpkgs: f24e81fbd3
- Hackage: 01c9e56f5d
- LTS Haskell: 87e2d54643
- Stackage Nightly: c30758374f
Adding stdenv.cc into the PATH, also setting CC, so that on Darwin
clang will be used by default. Still allowing to use an existing value
of CC if it is set already.
Replacing __inline_isnanl with __inline_isnan on darwin since the former
one was not defined.
This update was generated by hackage2nix v20151217-5-ged07a8e using the following inputs:
- Nixpkgs: 665c16fbe1
- Hackage: d9755ca900
- LTS Haskell: d3e5ae70f9
- Stackage Nightly: 56e1f693d5
Upstream bug: https://bugzilla.redhat.com/show_bug.cgi?id=1293060
This patch is based on the one attached to that bug report, but
instead of patching the .x files (parsing of which apparently
fails as well) it modifies the pre-generated .c files directly.
This ought to fix#12139.
Built and run locally.
From the Changelog:
```
Version 0.7.81, 2015-12-31
+ Acquisition Metadata: support of all SMPTE RDD18 elements
+ Matroska: cover presence and content of the cover, thanks to Max Pozdeev
+ #F446, Matroska: Handling of cropping values, thanks to Max Pozdeev
+ Improvement of Python binding: Mac Os X support, Python2 and Python3
can use same MediaInfoDLL.py
+ #F484, AVI: OpenDML Interlaced / Progressive scan type detection
+ MP4: support of AtomicParsley imdb tag
x #B959, MPEG-TS: MPEG-1 Video appeared as MPEG-2 Video
x #B914, Matroska: Undefined number of chapters in some M4V with Timed
Text, thanks to Max Pozdeev
x #B962, Matroska: negative timecodes were not correctly handled
x #B964, FLV: was hanging trying to open some FLV files
x JPEG in AVI or MOV: better handling of buggy APP0/AVI1, avoiding some
false positives about interlacement
x DVCPRO HD: some containers consider DVCPRO HD as with width 1920
despite the fact it is 1280 or 1440, using 1280 or 1440 in all cases
```
This update was generated by hackage2nix v20151217-3-gd4ae18a using the following inputs:
- Nixpkgs: 579e6bb797
- Hackage: e1530f9f9a
- LTS Haskell: d3e5ae70f9
- Stackage Nightly: db90cf927d
http://hydra.nixos.org/eval/1234895
The mass errors on Hydra seem transient; I verified ghc on i686-linux.
Only darwin jobs are queued ATM. There's a libpng security update
included in this merge, so I don't want to wait too long.
This improves our Bundler integration (i.e. `bundlerEnv`).
Before describing the implementation differences, I'd like to point a
breaking change: buildRubyGem now expects `gemName` and `version` as
arguments, rather than a `name` attribute in the form of
"<gem-name>-<version>".
Now for the differences in implementation.
The previous implementation installed all gems at once in a single
derivation. This was made possible by using a set of monkey-patches to
prevent Bundler from downloading gems impurely, and to help Bundler
find and activate all required gems prior to installation. This had
several downsides:
* The patches were really hard to understand, and required subtle
interaction with the rest of the build environment.
* A single install failure would cause the entire derivation to fail.
The new implementation takes a different approach: we install gems into
separate derivations, and then present Bundler with a symlink forest
thereof. This has a couple benefits over the existing approach:
* Fewer patches are required, with less interplay with the rest of the
build environment.
* Changes to one gem no longer cause a rebuild of the entire dependency
graph.
* Builds take 20% less time (using gitlab as a reference).
It's unfortunate that we still have to muck with Bundler's internals,
though it's unavoidable with the way that Bundler is currently designed.
There are a number improvements that could be made in Bundler that would
simplify our packaging story:
* Bundler requires all installed gems reside within the same prefix
(GEM_HOME), unlike RubyGems which allows for multiple prefixes to
be specified through GEM_PATH. It would be ideal if Bundler allowed
for packages to be installed and sourced from multiple prefixes.
* Bundler installs git sources very differently from how RubyGems
installs gem packages, and, unlike RubyGems, it doesn't provide a
public interface (CLI or programmatic) to guide the installation of a
single gem. We are presented with the options of either
reimplementing a considerable portion Bundler, or patch and use parts
of its internals; I choose the latter. Ideally, there would be a way
to install gems from git sources in a manner similar to how we drive
`gem` to install gem packages.
* When a bundled program is executed (via `bundle exec` or a
binstub that does `require 'bundler/setup'`), the setup process reads
the Gemfile.lock, activates the dependencies, re-serializes the lock
file it read earlier, and then attempts to overwrite the Gemfile.lock
if the contents aren't bit-identical. I think the reasoning is that
by merely running an application with a newer version of Bundler, you'll
automatically keep the Gemfile.lock up-to-date with any changes in the
format. Unfortunately, that doesn't play well with any form of
packaging, because bundler will immediately cause the application to
abort when it attempts to write to the read-only Gemfile.lock in the
store. We work around this by normalizing the Gemfile.lock with the
version of Bundler that we'll use at runtime before we copy it into
the store. This feels fragile, but it's the best we can do without
changes upstream, or resorting to more delicate hacks.
With all of the challenges in using Bundler, one might wonder why we
can't just cut Bundler out of the picture and use RubyGems. After all,
Nix provides most of the isolation that Bundler is used for anyway.
The problem, however, is that almost every Rails application calls
`Bundler::require` at startup (by way of the default project templates).
Because bundler will then, by default, `require` each gem listed in the
Gemfile, Rails applications are almost always written such that none of
the source files explicitly require their dependencies. That leaves us
with two options: support and use Bundler, or maintain massive patches
for every Rails application that we package.
Closes#8612
This update was generated by hackage2nix v20151217-3-gd4ae18a using the following inputs:
- Nixpkgs: 236677809b
- Hackage: 62b6b580fb
- LTS Haskell: d3e5ae70f9
- Stackage Nightly: ef03f60b53
Previously, the native libvirt package was making an assertion that
the dependent Python package had a compatible version. This commit
switches that so that the Python package makes the assertion, since
it makes more sense to me to have a child package making an
assertion about its parent than vice versa.
Previously the gems defaulted to "ruby" as the name and
"${ruby-version}-${gem-name}-${gem-version}" as the version,
which was just insane.
https://github.com/NixOS/nixpkgs/issues/9771#issuecomment-141041414
Noone is reacting so it's high time to take at least some action.
/cc @cstrahan.
This update was generated by hackage2nix v20151217-2-g519e0a9 using the following inputs:
- Nixpkgs: 2ed14d2303
- Hackage: 53c15ee9e3
- LTS Haskell: b668f53a86
- Stackage Nightly: ec9500b675
It is better to specify data-dir in the environmental variable since
then both the language description files and the dictionaries will be
found. Since dict-dir defaults to data-dir only the latter needs to be
set. See for example https://github.com/NixOS/nixpkgs/issues/1000
vcunat did some cosmetic changes, such as joining lines
because we seem to rarely use one-identifier-per-line style,
or fixing hyena description to conform to our rules.
Compatible with llvm+clang 3.7. Changes:
- Added Boost and Qt mappings.
- Better support for using declarations.
- Allow size_t from multiple headers.
- Fixed handling includes with common path prefix.
More: http://include-what-you-use.org/
This adds erlangPackages.${name}.env, which is usable to launch a shell
with the package and it's dependencies like this:
nix-shell -A erlangPackages.lager.env
This required:
1) refactoring buildHex to become a fixed-point function,
2) moves output of a package into $out/${name},
3) introduces erlEnv (buildEnv, which links in package and it's deps
into one directory - a super-simple plagiarization of
haskellPackages.compiler.ghcWithPackages) and
4) shell function producing a un-buildable derivation to be used by
nix-shell
This update was generated by hackage2nix v20151217 using the following inputs:
- Nixpkgs: 1e4c0752db
- Hackage: 42db5021ee
- LTS Haskell: 253d4da342
- Stackage Nightly: 57c8505aea
- The patch fixes building against gst-1.6.
- Having to change three files with almost same contents would drive me mad,
so I unified them into a single expression. /cc @ttuegel
- libxslt seemed unneeded, and it uses libxml2 anyway.
