can't distinguish between 5.10 and 5.14 packages. (We should
probably include the Perl version number in the name of Perl
packages, like GHC packages.)
svn path=/nixpkgs/trunk/; revision=33422
Markus Mottl <markus.mottl@gmail.com> said by email to me that
``Note that this link is served by a Mercurial Apache module, i.e. this
is not static content. It's likely that e.g. tar or bzip2 produce
slightly different output, which would explain changing hashes.
So I guess we shouldn't worry too much about the change in hash?
svn path=/nixpkgs/trunk/; revision=33419
The kmod tools appear to take a flag to name the location of the modules
directory, so we may not need to migrate our module-init-tools patches
svn path=/nixpkgs/trunk/; revision=33413
- cabal2nix: updated to version 1.30
- cmdargs: updated to version 0.9.5
- derive: updated to version 2.5.6
- hlint: updated to version 1.8.25
- io-choice: added version 0.0.1
svn path=/nixpkgs/trunk/; revision=33411
This expression can now be generated by cabal2nix. Note that it doesn't
propagate Emacs and haskell-mode, though, these dependencies must be
installed by the user in addition to ghc-mod.
To active ghc-mod, add the following snippet to your ~/.emacs file:
(autoload 'ghc-init "ghc" nil t)
(add-hook 'haskell-mode-hook (lambda () (ghc-init)))
svn path=/nixpkgs/trunk/; revision=33410
Linking these libraries makes sure that they can be found at run-time,
because a proper rpath into the Nix store is added to the generated
executable. Without that rpath, nbd-server will try to load the system's
libpthread.so.0, which may not be what we want.
svn path=/nixpkgs/trunk/; revision=33408
OTH, I'd prefer naming the attributes like "midoriWrapped" instead of "midoriWrapper".
But I chose the 2nd to follow the example of those already in all-packages.
svn path=/nixpkgs/trunk/; revision=33406
- LambdaHack: updated to version 0.2.1
- OpenAL: updated to version 1.4.0.1
- cabal2nix: updated to version 1.29
- unbound: updated to version 0.4.0.1
svn path=/nixpkgs/trunk/; revision=33394
By now, it happened twice that a commit broke GHC and thus all Haskell packages
we have in Nixpkgs. On such an occasion, I receive well in excess of 3000
notification e-mails from Hydra, and then I receive another 3000 e-mails after
the bug has been fixed. Under these circumstances, subscribing to these
notifications makes no sense for me.
svn path=/nixpkgs/trunk/; revision=33392
- cautious-file: updated to version 1.0.1
- hmatrix: updated to version 0.14.0.1
- random-fu: updated to version 0.2.2.0
- resourcet: added version 0.3.0
- sendfile: updated to version 0.7.6
- test-framework: added version 0.6
svn path=/nixpkgs/trunk/; revision=33354
