A full wrap isn't needed as these scripts are only run from within
hplip's own environment, and unwanted because they are imported
as Python modules.
Also fix a reference to /etc in $out/etc/hp/hplip.conf, though I've no
idea whether it is actually used.
Something changed in the generic fixupPhase magic exposing a bug
in the previous fix (wrapPythonPrograms where I should have used
wrapPythonProgramsIn).
Also move buildInputs and friends to a more conventional spot.
See issue #8694. Previously wombat256 was downloading its source from
the vim website using fetchurl. The builder was failing when it tried to
unpack the .vim source file. Using the vim-scripts github repo avoids
this problem.
The upstream tarball has changed, so I checked why this has happened and
found a tarball with the old SHA256 hash here:
http://fossies.org/linux/misc/xscreensaver-5.33.tar.gz/
After checking the contents of this and the new upstream tarball I found
that the old tarball had only .so, .o and .o.d files which now are no
longer existent in the new tarball.
Seems that the upstream author has accidentally put object files in the
source tarball and has now corrected this mistake.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
With this patch in place, ARMv7 NixOS can be booted in QEMU:
qemu-system-arm -kernel u-boot -M vexpress-a9 -serial stdio -sd nixos-sd-image-armv7l-linux.img -m 1024
...with all the features that boot.loader.generic-extlinux-compatible
supports, like the boot generation menu and seamless kernel upgrades
in the VM.
Instead of selecting the defconfig based on stdenv.platform.uboot,
provide different ubootFoo packages. Otherwise we couldn't easily build
U-Boots for different platforms than what we are currently running on.
All users of the ubootChooser function appear to be using only CLI tools
like mkimage, whose behaviour is not affected by the defconfig (their
build outputs are bitwise-identical). So add a separate package for the
CLI tools.
Of the removed patches, some version of sheevaplug-sdio.patch has
apparently been applied upstream (with at least mv_sdio.c renamed to
mvebu_mmc.c). sheevaplug-config.patch needs rebasing & re-testing on
real hardware.
Tested boards and input/output methods that upstream supports:
- Raspberry Pi:
- HDMI works, USB keyboard not yet supported
- Serial via the 26-pin connector (3.3V)
- pcDuino3 Nano:
- HDMI + USB keyboard (only if attached to a hub)
- Serial via the 3-pin connector (3.3V)
- Jetson TK1: RS-232 serial port only
- Versatile Express CA9 (for QEMU only): Serial via '-serial stdio'
- upgrade wakatime cli to v4.0.14
- make sure config file has api_key
- only display setup complete message first time setting up cfg file
- don't log time towards git temporary files
- prevent slowness in quickfix window to fix#24
- reuse SSL connection across multiple processes for improved performance
- correctly display caller and lineno in log file when debug is true
- project passed with --project argument will always be used
- new --alternate-project argument
- fix bug with auto detecting project name
- correctly log message from py.warnings module
- handle plugin_directory containing spaces
None of the installed tools like hp-align or hp-toolbox could
work, because they tried to import wrapped Python scripts (i.e.
shell code) as Python modules. Fix this by:
- Substituting `dbus` for the defective `pythonDBus`.
- No longer blindly wrapping all python files, but replacing
the original symlinks in $out/bin with wrapper scripts.
These wrappers then `exec` the unmodified Python scripts in
$out/share/hplip directly, instead of the .<name>-wrapped
copy that wrapPythonProgramsIn creates. The latter are simply
removed.
Sure, it's a bit ugly. It's also 2015 and I couldn't use my
printer :-/
Now, the expression will not install ROM files anymore. It is easier the
final user install them locally by himself, in a local directory.
Also, it appears that there is a missing ROM in the package provided by
the project's site.
This not the latest, but the greatest version (by number of models still
supported).
A big bump. Everything seems to work fine. Tested with a Samsung CLP-325.
Also fixed some strange things. Did this ever even work? On x86_64?
Foomatic filters contained a 64-char c string hardcoded to /bin/bash.
This caused some filters (at least pdftops) to fail.
I also had to increase the size of the string because nix paths are too
long.
It was trying to find "gs" via execve, so use execvpe instead. It's
probably better to use gs's absolute path, but maybe not every
cups-filters user needs it.
- sort automatcially generated vim derivations
- move the plugins to be derived by VAM into its own file: vim-plugin-names
- rename vimrc into vimUtils moving all common code into it
so that it is possible to use it within ~/.nixpkgs/config.nix
- Moving all important documentation into vimUtils
- provide vimPlugins.pluginnames2Nix to provide a bin/vim-* command
writing updated derivations into its buffer
VAM's repository's nix.vim got some improvements @ VAM repository, eg hg
fetchout out got fixed and shell failures are noticed now (Thanks to Arseniy
Seroka)
Tons of accuracy improvements per usual, some somewhat interesting ones:
- Constant Angular Velocity for Disk Reads
- FIFO buffer fix
- Vectorized Vertex loading
and more! read about it at dolphin-emu.org
initialPath is now added to the top of the modified setup.sh. It's not
very nice altogether to patch setup.sh in this way, as it is prone to
break on setup.sh changes, but it should (hopefully finally, but
unfortunately only temporarily) fix the issue ewemoa (in #nixos) had
with myEnvFun.
In the long term, if people want to have roughly similar functionality,
nix-shell provides a better way for that, so maybe we might either
remove myEnvFun or replace it with a smaller implementation using
nix-shell.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
If you do not provide any dependencies which don't propagate coreutils,
you'll get "rm: No such file or directory". Like for example by using
this:
nix-build -E '(import <nixpkgs> {}).myEnvFun {
name = "mybrokentest";
buildInputs = [];
}'
Thanks to "ewemoa" from the #nixos IRC channel for spotting this.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
This introduces Chromium 39 as the new stable version along with a bunch
of fixes.
Fixes#2799, particularily the PDF plugin, which now is open source and
thus no longer an issue.
Also fixes#3219 and merges #2906, so we no longer get a crash while
trying to bring up the print preview dialog.
Thanks to @edwtjo for the CUPS version bump.
* chromium: Switch to use open-source PDF plugin.
* cups: bump 1.5.4 -> 1.7.5
* chromium: Allow env vars for passing plugin paths.
* chromium: Update all channels to latest versions.
* protobuf: Clean up and update to version 2.6.1.
This reverts commit 1b064645ed, reversing
changes made to 14cef95e84.
As discussed in the comments to the aforementioned pull request, while
the new version supports many new printer models, many components of the
driver are broken because of some corner cases in the Python wrapper. In
particular, the loader for the binary binary driver needed by many
middle- and high-end printers is broken, negating most of the value in
updating!
This patch implements derving a .vimrc from vim-plugins.nix loading those
plugins by either Pathogen or VAM (VAM seems to be slightly faster and is much
more powerful).
Example:
environment.systemPackages = [
# default plain vim
vim_configurable
# vim which get's called vim-with-addon-nix
(vim_configurable.customize {
name = "vim-with-addon-nix";
vimrcConfig.vam.pluginDictionaries = [{name = "vim-addon-nix"; }];
})
];
This way you can provide an "enhanced Vim" and a standard Vim.
Details about what this commit changes:
1) provide a new toplevel name vimrc which
* provides a way to build up a .vimrc using either pathogen or VAM (knowing about plugin dependencies by name)
* can enhance vim to support. vim.customize { name = "name-user"; vam.pluginDictionaries and/or pathogen.pluginNames = .. }
* introduce rtp names for each vim plugin pointing to the runtimepath path
* suggest naming to be the same as vim-pi so that VAM's dependencies work
* derive some packages as example from vim-pi using VAM's new autoload/nix.vim
supporting simple dependencies
* test case for vim-addon-nix for VAM/pathogen
2) enhance vim_configurable to support .customize
3) update many plugins by using VAM's implementation not rewriting those which
* vim-pi doesn't know about the git source yet (TODO: make vim-pi be aware of
those)
* have special build code
This commit partially conflicts with commits done by Bjørn Forsman starting by
37f961628b, eg the one using lower case attr and pkg names, because they don't
match vim-pi (eg YouCompleteMe). Rather than resolving the conflict this just
adds aliases so that both names can be used
Git commit ids shouldn't be used as version numbers in nixpkgs:
* they don't increase monotonically
* they don't always start with a digit, causing nix to not parse them as
version numbers, and instead sees them as part of the package name,
causing the package to (seemingly) conflict with itself upon upgrade.
And use fetchzip instead of fetchurl towards github.com to not let this
happen again. (fetchzip checks the hash on the *extracted* archive, so
changes in compression algo etc. doesn't change the hash.)
Currently some plugins use "vim-" prefix, some use "-vim" suffix and
some have no prefix or suffix at all.
I chose "vimplugin-" prefix to match the attribute path: vimPlugins.*
retroarchBare was too old for any practical use at this point, so it was removed in favor
of fetching releases from git. RetroArch often recommends checkout out the latest git master
if you are on a desktop platform, so we are conforming with their recommendation. I plan on
bumping the git of retroarch and each core every month, and I plan on adding more libretro
cores in the future.
The libretro cores that were added are nestopia, quicknes, snes9x, and vba-m
planned next are : bsnes-accuracy, bsnes-balanced, bsnes-performance, a few libretro mame ports
Closes#4448
Using a git commit id as version number doesn't work well, so use the
commit date instead.
Also, there is no other version of YouCompleteMe than the git version,
so drop 'git' from the package name.
patchShebangs fixes build failure in chroot (in this new version).
The --system-libclang flag is now needed to prevent YouCompleteMe from
trying to download and build clang (which fails).
These are wrappers for retroarch wrappers, they are needed since launching a
process from XBMC(as a display manager and probably otherwise), using
AdvancedLauncher, results in it and it's parent recieving the same gamepad
input.
These wrappers will produce no sound on XBMC and AdvancedLauncher setups not
using a sound daemon, since XBMC gets paused while still holding onto the sound
device.
This reverts commit 13fab57d12.
Hopefully fixes the printing tests at
http://hydra.nixos.org/build/14952070/log/raw ; if not, it's probably
gutenprint although I doubt:
client# lp: Unsupported document-format "application/pdf".
Now we can parametrize Higan locally.
By default, guiToolkit = "gtk" and profile = "performance" (the accuracy
profile is seriously slow on my machine :) )
Closes#4340
Sound of Sorting is an array-sorting visual+sound demo program.
It shows an array as a list of horizontal bars, and realizes a
step-by-step sorting of it. Moreover, it colorizes and emits a
"8-bit-game-like" sound throughout its execution.
Closes#4341