Built and tested locally.
From the changelog:
```
Version 0.7.84, 2016-03-31
x #I122, MPEG-4/MOV: Crash if mdhd timescale is 0
x MPEG-4/MOV: Infinite loop if malformed stsc / stsz (detected by
fuzzing)
x MPEG-TS: some DVB Subtitles were not detected
x HLS: better handling of media playlists having EXT-X-BYTERANGE
```
Overview of the updated versions:
stable: 49.0.2623.87 -> 49.0.2623.110
beta: 50.0.2661.26 -> 50.0.2661.49
dev: 50.0.2661.18 -> 51.0.2693.2
Most notably, this includes a series of urgent security fixes:
* CVE-2016-1646: Out-of-bounds read in V8. Credit to Wen Xu from
Tencent KeenLab.
* CVE-2016-1647: Use-after-free in Navigation. Credit to anonymous.
* CVE-2016-1648: Use-after-free in Extensions. Credit to anonymous.
* CVE-2016-1649: Buffer overflow in libANGLE. Credit to lokihardt
working with HP's Zero Day Initiative / Pwn2Own.
* CVE-2016-1650: Denial of service in PageCaptureSaveAsMHTMLFunction
The official release announcement with details about these fixes can be
found here:
http://googlechromereleases.blogspot.de/2016/03/stable-channel-update_24.html
Beta and stable could be also affected, although I didn't do a detailed
check whether that's the case.
As this introduces Chromium 51 as the dev version, I had to make the
following changes to make it build:
* libexif got removed, so let's do that on our end as well.
See https://codereview.chromium.org/1803883002 for details.
* Chromium doesn't seem to compile with our version of libpng, so let's
resort to the bundled libpng for now.
* site_engagement_ui.cc uses isnan outside of std namespace, so
we're fixing that in postPatch using sed.
I have successfully built all versions on i686-linux and x86_64-linux
and tested it using the VM tests.
Test reports can be found at the following evaluation of my Hydra:
https://headcounter.org/hydra/eval/314584
Thanks to @grahamc for reporting this.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
Reported-by: Graham Christensen <graham@grahamc.com>
Fixes: #14299
We were using a version years old.
I picked the git version instead of the last release because the git version
has fixes not released for almost one year.
I updated the erlang esdl lib, added the erlang cl lib,
added opencl-headers and ocl-icd to make wings build and run.
I have not tested its opencl part; I only added dependencies so
it builds.
I originally wanted to do this a long time (a31301d) but IIRC back then
it didn't compile. Nowadays with the splitup of the gold linking flags
and the binutils integration, it's merely just a switch to flip, so
let's do that.
Only tested it by building against the current Chromium stable version
on 64bit, because right now builds on Hydra seem to time out (because of
this?) anyway so we have nothing to lose here.
The linking time was hereby reduced from >30 minutes (I didn't measure
it exactly but looked half an hour later to the build progress and it
was *still* linking) to about a few seconds, which I guess is even
though the measurement is quite bogus a tremendous improvement
nonetheless.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
The openimageio dependency doesn't build on i686. But probably nobody
cares about running Blender on 32-bit anymore.
http://hydra.nixos.org/build/33602734
Note: the 1.2.9 GitHub tarball's sha256 has changed.
Changes:
- Fixed an issue introduced by the fix for 1196 that had the "Upload
to SD" button stop working correctly.
- Fixed an issue causing an error on disconnect after or cancelling of
an SD print, caused by the unsuccessful attempt to record print
recovery data for the file on the printer's SD card.
- Only add bed temperature line to temperature management specific
start gcode in CuraEngine invocation if a bed temperature is actually
set in the slicing profile.
- If a communication timeout occurs during an active resend request,
OctoPrint will now not send an M105 with an increased line number
anymore but repeat the last resent command instead.
- Don't add an extra ok for M28 response.
- Add an extra ok for M29 response, but only if configured such in
"Settings" > "Serial" > "Advanced options" > "Generate additional ok
for M29"
- Trigger M20 only once after finishing uploading to SD
- Prevent M105 "cascade" due to communication timeouts
- Fixed wrong tracking of extruder heating up for M109 Tn commands in
multi-extruder setups.
- Fixed start of SD file uploads not sending an M110.
- Fixed job data not being reset when disconnecting while printing.
Commit aa097946d2 only fixed evaluation.
Ssince 37dbd62 however, the fetchurl call is already implied so just
changing the path will still result in fetchurl (fetchurl ...), so let's
drop the outer fetchurl.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
Cc: @msteen, @benley
Fixes#12794 by reverting the source tree splitup (c92dbff) to use the
source tarball directly into the main Chromium derivation and making the
whole source/ subdirectory obsolete. The reasons for this are explained
in 4f981b4f84.
This also now renames the "sources.nix" file to "upstream-info.nix",
which is a more proper name for the file, because it not only contains
"source code" but also the Chrome binaries needed for the proprietary
plugins (of course "source" could also mean "where to get it", but I
wanted to avoid this ambiguity entirely).
I have successfully built and tested this using the VM tests.
All results can be found here:
https://headcounter.org/hydra/eval/313435
As of 6041cfe, the upstream-info.nix (back then it was called
sources.nix) is no longer in the source/ subdirectory, so we need to fix
that comment to say that the file is autogenerated from update.sh in the
*same* directory.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
This reverts commit 5979946c41.
I have tested this by building against the stable version of Chromium
and it seems to compile just fine, so it doesn't seem to be needed
anymore.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
* the major change is to set TARGETDIR=${vardir}, and symlink from
${vardir} back to ${out} instead of the other way around. this
gives CP more liberty to write to more directories -- in particular
it seems to want to write some configuration files outside of conf?
* run.conf does not need 'export'
* minor tweaks to CrashPlanDesktop.patch
Only a aesthetics thingy, but also corrects the comment, because we're
essentially precompiling .py files, NOT the .pyc files (the latter are
the results).
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
This addresses #12794 so that we now have only a single tarball where we
base our build on instead of splitting the source into different outputs
first and then reference the outputs.
The reason I did this in the first place is that we previously built the
sandbox as a different derivation and unpacking the whole source tree
just for building the sandbox was a bit too much.
As we now have namespaces sandbox built in by default we no longer have
that derivation anymore. It still might come up however if we want to
build NaCl as a separate derivation (see #8560), but splitting the
source code into things only NaCl might require is already too much work
and doesn't weight out the benefits.
Another issue with the source splitup is that Hydra now has an output
limit for non-fixed-output derivations which we're already hitting.
Tested the build against the stable channel and it went well, but I
haven't tested running the browser.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
We always do something like "fetchurl channelProduct", so let's move it
to getChannel directly so we can avoid those fetchurl calls all over the
place.
Also, we can still access subattributes from the fetchurl call if we
need to, so there really is no need to expose the product's attributes
directly.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
Yes, I know I'm a bit nitpicky, but lines >80 chars are very ugly if you
have two windows side-by-side.
Thus no feature changes here.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
The plugin was looking for pvr.hts.so.2.2.13, so as only pvr.hts.so was copied, the plugin would not load. This symlinks all variants of the name to the addons folder.
We now should have only the default.nix left in the source directory and
we can start to factor out the pieces into the Chromium main derivation
attributes.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
The "sources.nix" also contains information about where to get binary
packages, so calling it "upstream-info.nix" fits better in terms of
naming.
Also, we're moving it away from the sources dir, because the latter will
soon vanish.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
We're going to reference the patches in the Chromium main build rather
than applying it to the sources. So as a first step, this should keep
the patches away from the "source" subdirectory so we can make it flat.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
We were previously getting collissions for the following
2 files with files from the dropbox package:
- bin/dropbox
- share/applications/dropbox.desktop
As a consequence the binary has been renamed to dropbox-cli and
the .desktop removed as it is redundant.
Five years in the making. Fixed bugs:
* Fix LASH support (crash on 64 Bit systems)
* Fix broken JACK transport with newer jackd version
* Fix clock tick drift
* Fix jack session commandline (obsolete --file option removed)
General Changes:
* C++11 compatible compiler required
Built and tested locally.
From the changelog:
```
Firefox 45/46 compatibility
* [Zotero Standalone] Save PDFs directly from browsers via the Save to
Zotero button
* Requires Chrome connector version 4.0.29; updated Safari connector
coming soon
* To convert a standalone PDF attachment to a full item, right-click
the item and choose “Retrieve Metadata for PDF” or “Create Parent
Item”.
* [Zotero for Firefox] Don't show Firefox open/save dialog when
showing style installation prompt
* Retry additional file sync failures automatically
* Fixed potential crash when dragging in files on some systems
* Fixed potential crash when dragging collections to another library
* Miscellaneous bibliographic output and word processor integration
fixes
* Updated citeproc-js to version 1.1.71
* Fixed spacing of toolbar icons on Windows/Linux in Firefox 43+
* Use larger text field for Book Title
* Make highlight color of selected items list rows on Linux more
distinct when items pane doesn't have focus
* [Zotero for Firefox] Improved error message on Firefox login manager
failure
* Improved error message when trying to use a later database version
```
Building the manual requires docbook_xsl; in sandboxed environments,
the build would fail due to being unable to download missing docbook
resources.
Also include some minor improvements to the build recipe:
- use fetchFromGitHub
- move build-time dependencies to nativeBuildInputs
xref: https://github.com/NixOS/nixpkgs/pull/13900
This is just a minor upgrade, even though the commit message says it's
to major version 50. However, the CVEs listed there are for real, see
the following announcement:
http://googlechromereleases.blogspot.de/2016/03/stable-channel-update_8.html
The summary of updated packages:
stable: 49.0.2623.75 -> 49.0.2623.87
beta: 49.0.2623.75 -> 50.0.2661.26
dev: 50.0.2661.11 -> 50.0.2661.18
I've also added two commits, fixing the chdir() in the updater and
shutting up Python precompilation errors during the preBuild phase.
Tested on my Hydra at:
https://headcounter.org/hydra/eval/312166
Changing the working directory to
pkgs/applications/networking/browsers/chromium is a bit annoying, so
let's make sure the script can be called from anywhere.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
The errors are completely non-fatal and only cause a particular file to
be not precompiled. Unfortunately this can lead to confusion to whether
these errors are real errors or not, so let's shut it up completely
because they're *not* real errors.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
As of version 2.92, transmission-cli is no longer built by default (it
is deprecated). This breaks the bittorrent vmtest. For now, explicitly
enable the cli.
Built and tested locally.
From the changelog:
```
Version 0.7.83, 2016-02-29
+ HEVC: Maximum Content Light Level (MaxCLL) and Maximum Frame-Average
Light Level (MaxFALL), metadata mandated by CEA-861.3 for HDR support
+ HEVC: Mastering display color primaries and luminance (based on SMPTE
ST 2084), metadata mandated by CEA-861.3 for HDR support
+ HEVC: SMPTE ST 2048 and SMPTE ST 428-1 transfer characteristics
+ HEVC: Chroma subsampling location (indication of the location type
described in the HEVC spec)
+ MPEG-TS: ATSC Modulation Mode (Analog, SCTE_mode_1 aka 64-QAM,
SCTE_mode_2 aka 256-QAM, 8-VSB, 16-VSB)
+ #B981, MP4: support of buggy file having "hint" SubType
x HLS: better handling of media playlists having EXT-X-BYTERANGE
```
Additionally, some cleanup and package uses `autoreconfHook` now. Thanks
to @hrdinka for helpful pointers!
Commit 4a54794d18 upgraded Thunderbird's
version to 38.6.0 (accidentally?), but didn't change the hash. This
wasn't caught due to tarballs.nixos.org being keyed on hash only.
Fixes https://github.com/NixOS/nixpkgs/issues/12749
The build failure was caused by brittle detection of the
llvm version. See the code for (excessive) details. This fix
is a quick hack, a proper fix would be to parse the version
of the input llvm derivation and use that to derive a proper
value. Here we just pin the version.
Also move build-time deps to `nativeBuildInputs`.
rq only compiles with ruby 1.8 which we don't distribute anymore.
the source is dead.
there is a 1.9 branch over https://github.com/pjotrp/rq that hasn't been
touched for 4 years.
All the programs provided by ncurses were being installed to the $dev
output, but several of them are intended for runtime use, e.g. to
operate on the running terminal. These user-facing programs are moved to
the $bin output.
Several packages referred to "${ncurses}/bin" or "${ncurses.dev}/bin" at
runtime; these paths are also updated to refer to "${ncurses.bin}/bin".
The $lib output refers to the terminfo database in $out, which is about
10x larger than the ncurses shared library. Splitting these outputs
saves a small amount of space for any derivations that use the terminfo
database but not the ncurses library, but we do not have evidence that
any such exist.
- Now usable in non kde desktop environments.
Build a immutable sycoca database and use
wrappers to tie programs to this
database and avoid interference from the
outside by specifying a fixed `KDELIBS`
and fixed/empty `XDG_DATA_DIRS`.
Added missing dependencies to syscoca
database so that the program is complete.
Added all build time optional packages.
Kipi-plugins now properly detected. Added
almost all optional dependencies so that
almost all plugins are usable.
Now with vlc phonon backend for video playback.
Now with ffmpeg thumbnailer for video items
thumbnail creation.
Now run without any error log.
Tests:
- Ran most features of the standard program. Everything
work perfectly without error logs.
- Ran some of the kipi plugins. Work fine there too.
- Ran face detection and fingerprint generation
successfully.
- Oxygen icons are now displayed properly.
- Ran other wrapped executable successfully.
Overview of the updated versions:
stable: 48.0.2564.116 -> 49.0.2623.75
beta: 49.0.2623.63 -> 49.0.2623.75
dev: 50.0.2657.0 -> 50.0.2661.11
Stable and beta are now in par because of the release of a major stable
update.
The release addresses 26 security vulnerabilities, the following with an
assigned CVE:
* CVE-2016-1630: Same-origin bypass in Blink. Credit to Mariusz
Mlynski.
* CVE-2016-1631: Same-origin bypass in Pepper Plugin. Credit to Mariusz
Mlynski.
* CVE-2016-1632: Bad cast in Extensions. Credit to anonymous.
* CVE-2016-1633: Use-after-free in Blink. Credit to cloudfuzzer.
* CVE-2016-1634: Use-after-free in Blink. Credit to cloudfuzzer.
* CVE-2016-1635: Use-after-free in Blink. Credit to Rob Wu.
* CVE-2016-1636: SRI Validation Bypass. Credit to Ryan Lester and
Bryant Zadegan.
* CVE-2015-8126: Out-of-bounds access in libpng. Credit to
joerg.bornemann.
* CVE-2016-1637: Information Leak in Skia. Credit to Keve Nagy.
* CVE-2016-1638: WebAPI Bypass. Credit to Rob Wu.
* CVE-2016-1639: Use-after-free in WebRTC. Credit to Khalil Zhani.
* CVE-2016-1640: Origin confusion in Extensions UI. Credit to Luan
Herrera.
* CVE-2016-1641: Use-after-free in Favicon. Credit to Atte Kettunen of
OUSPG.
The full announcement which also includes the link to the bug tracker
can be found here:
http://googlechromereleases.blogspot.de/2016/03/stable-channel-update.html
Also, the 32bit Chrome package needed for the Flash and Widevine plugins
doesn't exist anymore, because Google has dropped support for 32bit
distros, see here for the announcement:
https://groups.google.com/a/chromium.org/forum/#!topic/chromium-dev/FoE6sL-p6oU
On our end, we need to fix the patch for the plugin paths to work for
the latest dev channel. The change is very minor, because the
nix_plugin_paths_46.patch only doesn't apply because of an iOS-related
ifdef.
Built and tested on my Hydra at:
https://headcounter.org/hydra/eval/311511
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
Fixes: #13665
Comparing the current version with the version in sources list and
accidentally swapping the version arguments isn't going to get very far
because every new version that will come up will then be treated as "we
already have that version".
So we're now using versionOlder and also a check whether the version is
the *same* as the one in sources.nix.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
fixes#13651
@edolstra:
“IIRC, the "KDE integration" consists mainly in providing a Qt/KDE "File
open" dialog, which we can probably live without. In fact, it has been
known to malfunction if Libreoffice's KDE version is different from the
system KDE version.
So disabling this sounds like a good idea.”
`*.desktop` file now:
- Refers to an icon.
- Is placed in the proper category (based on comparison with `keepassx2`.
- Has proper mime type (also based on comparison).
Also, now use `icoutils` to extract icons from the application ressouces
and transform them from `*.ico` to `*.png`. Created and used a generic
script that has the ability to move the extracted `*.png` to their
appropriate standard freedesktop location.
Tested this on nixos. `keepass` now has a icon and is categorized
in the same bin as `keepassx2`. The program still execute and
function prefectly.
- Replace non freedesktop `CAD` desktop category
by categories taken from debian's desktop item.
This make the desktop items appear under the
`Graphic` category in DE menus instead of `Others`.
- Moved to new package icons as specified by
upstream release notes.
- Moved to icon name instead of full path to
a single icon as specified by freedesktop
specification.
- Human readable desktop item names.
Updates VirtualBox from version 5.0.12 to 5.0.14.
Upstream changes are (without bug IDs):
* GUI: properly limit the number of VCPUs to the number of physical cores
on Mac OS X
* Audio: fixed a bug which prevented loading a saved state of a saved
guests with HDA emulation (5.0.12 regression)
* Audio: don't crash if the backend is unable to initialize
* Audio: fixed audio capture on Mac OS X
* Storage: fixed a possible crash when attaching the same ISO image
multiple times to the same VM
* BIOS: properly report if two floppy drives are attached
* USB: fixed a problem with filters which would not capture the device
under certain circumstances (5.0.10 regression)
* ExtPack: black-list Extension Packs older than 4.3.30 due to
incompatible changes not being properly handled in the past
* Windows hosts: fixed a regression which caused robocopy to fail
* Linux hosts: properly create the /sbin/rcvboxdrv symbolic link (5.0.12
regression)
* Mac OS X hosts: several fixes for USB on El Capitan
* Linux Additions: fixes for Linux 4.5
Full upstream changelog with bug IDs can be found at:
https://www.virtualbox.org/wiki/Changelog
The reason I was reluctant to merge this before were these symbol lookup
errors:
vboxsf: Unknown symbol VBoxGuest_RTMemTmpFree (err 0)
vboxsf: Unknown symbol VBoxGuestIDCCall (err 0)
vboxsf: Unknown symbol VBoxGuest_RTSemFastMutexRequest (err 0)
vboxsf: Unknown symbol VBoxGuest_RTSemFastMutexRelease (err 0)
vboxsf: Unknown symbol VBoxGuest_RTLogRelGetDefaultInstanceEx (err 0)
vboxsf: Unknown symbol VBoxGuest_RTErrConvertToErrno (err 0)
vboxsf: Unknown symbol VBoxGuest_RTSemFastMutexCreate (err 0)
vboxsf: Unknown symbol VBoxGuest_RTSemFastMutexDestroy (err 0)
vboxsf: Unknown symbol VBoxGuest_RTMemContFree (err 0)
vboxsf: Unknown symbol VBoxGuest_RTSemMutexRelease (err 0)
vboxsf: Unknown symbol VBoxGuestIDCOpen (err 0)
vboxsf: Unknown symbol VBoxGuest_RTAssertShouldPanic (err 0)
vboxsf: Unknown symbol VBoxGuest_RTMemContAlloc (err 0)
vboxsf: Unknown symbol VBoxGuest_RTSemMutexRequest (err 0)
vboxsf: Unknown symbol VBoxGuest_RTSemMutexCreate (err 0)
vboxsf: Unknown symbol VBoxGuest_RTMemTmpAllocTag (err 0)
vboxsf: Unknown symbol VBoxGuest_RTSemMutexDestroy (err 0)
vboxsf: Unknown symbol VBoxGuest_RTAssertMsg1Weak (err 0)
vboxsf: Unknown symbol VBoxGuestIDCClose (err 0)
vboxsf: Unknown symbol VBoxGuest_RTAssertMsg2Weak (err 0)
However, after testing it against 5.0.12, the same errors occur there as
well, so it is likely related to our VM tests.
stdenv.system should be almost always used instead of builtins.currentSystem
or cross-evaluation (e.g. evaluating a i686 NixOS system on a 64-bit nix)
will be subtly broken.
This makes pythonPackages.sqlalchemy the most up to date revision (it
was called sqlalchemy_1_0 before), and maintains the various “legacy”
versions available as pythonPackages.sqlalchemyX for X in {7,8,9}.
All derivations that required `sqlalchemy_1_0` now require `sqlalchemy`
while those that required `sqlalchemy` now require `sqlalchemy7`.
The derivations are not changed, only the attribute names they are
bound to.
No changes in functionality, but to make future source updates a bit
easier on the eyes when viewing the diff.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
The update.sh shell script now is only a call to nix-build, which does
all the hard work of updating the Chromium source channels and the
plugins. It results in a store path with the new sources.nix that
replaces the already existing sources.nix.
Along the way, this has led to a quite massive workaround, which abuses
MD5 collisions to detect whether an URL is existing, because something
like builtins.tryEval (builtins.fetchurl url) unfortunately doesn't
work. Further explanations and implementation details are documented in
the actual implementation.
The drawback of this is that we don't have nice status messages anymore,
but on the upside we have a more robust generation of the sources.nix
file, which now also should work properly on missing upstream
sources/binaries.
This also makes it much easier to implement fetching non-GNU/Linux
versions of Chromium and we have all values from omahaproxy available as
an attribute set (see the csv2nix and channels attributes in the update
attribute).
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
This fixes the previous attempt (in commit 46d1dd5) to make ctags' ctags
override emacs' ctags. The higher the value of the priority attribute,
the higher the priority (see the definition of `buildEnv`).
