Notable changes:
* deps.qt: Add missing obs-deps version file for Qt packages
* CI: Add detection for single-arch static libraries
This is mostly to fix plugin builds. No dependency versions have
changed.
Notable changes:
* deps.ffmpeg: Update FFmpeg to a6dc92968a
* deps.qt: Backport macOS disabled combobox items fix
The FFmpeg update is to fix an issue with DTS when using b-frames with
FFmpeg NVENC. In particular, we wanted this commmit:
* https://github.com/FFmpeg/FFmpeg/commit/a6dc92968a325d331bb6dcf9b3b2248026cd1d6c
* x264: Updated to the same commit as obs-deps
* nv-codec-headers: Update to n12.0.16.0
* aom: Updated to 3.6.0
* svt-av1: Updated to 1.4.1
* ffmpeg: Updated to 6.0
* luajit: Updated to the same commit as obs-deps
* x264: Updated to the same commit as obs-deps
* srt: Updated to 1.5.1
* aom: Updated to 3.5.0
* svt-av1: Updated to 1.3.0
* ffmpeg: Updated to the same commit and patches as obs-deps
* luajit: Updated to the same commit as obs-deps
The 6.4 runtime rely on Freedesktop 22.08 which provide a more recent
version of Mesa and PipeWire.
The switch to 22.08 also requires to migrate paths of VST 2 plugins.
In dfccb26bb2, in my haste to fix Windows
CI, I accidentally included a change to disable Windows scripting which
I had been using locally. Reverse that change to re-enable scripting in
Windows builds.
The backtick double-quote pattern does not work in PowerShell 7.3.x.
There are still some other possible PowerShell 7.3.x issues in the
packaging steps, but let's fix this first to get regular builds working
in PowerShell 7.3.x again.
Apparently Steam running the OBS launch script makes it run
inside Rosetta as well, so it ended up launching the x86 version
even on arm64 systems.
Explicitly detect Rosetta translation and set arm64 in those cases.
The CI build script is useful to build OBS Studio even outside the CI
environment. However, the current script hard codes adding the 'amd64'
architecture to dpkg, the Debian/Ubuntu package management database.
This breaks the system when run on non-x86 machines. e.g., a
linux-aarch64 VM running on an Apple silicon macbook. Fixing the system
requires telling dpkg to remove 'amd64' again with 'sudo dpkg
--remove-architecture amd64'
As the CI build always runs on an amd64 machine, adding amd64 is not
necessary and can be safely removed.
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
Fixes#7146
Normally, Qt and CEF each run their own version of GTK, often causing
conflicts. This commit introduces a build of CEF where GTK has been
disabled using `use_gtk=false` as we do not require the Chrome runtime.
This is an experiment and may end up being the safest long-term solution.
Note: This does introduce potential quirks/incompatibilities, such as
browser sources no longer being able to display alert() dialogs, and
potentially file browser dialogs not working from CEF.
This is better than the alternative. If necessary, such dialogs can be
reimplemented using Qt at a later date, and would likely provide a more
consistent user experience.
On CI, do not fetch tags in packaging scripts. For some reason, the
checkout action seems to locally update any new git tags on the runner:
t [tag update] (commit-hash) -> tag-name
This causes future calls to fetch git tags to fail on CI with:
! [rejected] tag-name -> tag-name (would clobber existing tag)
To avoid this, we can simply not fetch tags a second time on CI.
Additionally, fix the Windows Installer job.
* x264: Updated to the same commit as obs-deps
* srt: Updated to 1.5.0
* mbedtls: Updated to 3.2.1
* librist: Updated to the same commit as obs-deps
* aom: Updated to 3.4.0
* svt-av1: Updated to 1.1.0
* ffmpeg: Updated to the same commit and patches as obs-deps
* luajit: Updated to the same commit as obs-deps
In .github/workflows/main.yml, for the linux_build job, the variable
BUILD_FOR_DISTRIBUTION is set to the string "true" or "false" on CI.
Later, in CI/linux/03_package_obs.sh, we perform a boolean check on
this variable. However, "false" will evaluate as true, because it is a
non-null string. This was causing CI Linux packages to always build as
if BUILD_FOR_DISTRIBUTION was enabled, which caused the git commit hash
to be omitted from package filenames.
Since we know the expected values, let's just test directly if the
variable equals "true" to get the expected behavior.
derrodandRyan Foster
gxalphaandJim
tytan652andGeorges Basile Stavracas Neto
PatTheMavandRyan Foster
Jan BeichandRyan Foster
Grant LikelyandJim
Matt Gajownik
cg2121andRyan Foster
tytan652andGeorges Basile Stavracas Neto
CarrotCodesandMatt Gajownik
Norihiro KamaeandPatrick Heyer