Extend the script added by previous commit to one handling the entire
bundling from Makefile.
It is nicer having the scripting outside the Makefile, anyways, and the
bundle LC_RPATH fix workaround closly relates to bundling.
Fixes the UG crashes caused by the multiplication of
"@executable_path/../libs/" caused by replacing multiple (originally
distinct) LC_RPATH values with this one.
+ removed invalid comment from Makefile (actually was related to
"bundle-nolib" target that is no longer present)
closes GH-436
- setup_firejail was not passed the whole command-line as the fucnction
(and subsequently called ones) expected
- omit adding --read-only=/tmp fixed - now paths are "-enclosed, handle
eventual mutliplied leading / and evental trailing one and don't assume
space after (not needed now, " delimits that)
If ULTRAGRID_AUTOUPDATE is set and the update is due, touch the AppImage
first. This is convenient if user wants to skip the update eg. by pressing
Ctrl-C, next run will not enforce the update becauuse the date of the file
is reset. User can run then run the AppImage by issuing the command again
without the update.
+ notice that updating because of ULTRAGRID_AUTOUPDATE is set
+ rename handle_updates to handle_autopudates (the normal update by -u is
not handled by that function at all)
If $APPIMAGE_DEBUG print explicitly that no updates are
available. (`appimageupdatetool -j` really returns 0 when no updates
available, otherwise 1 or a different value on error).
macOS 12 is no longer supported and as Homebrew has a policy, that it
compiles packages from sources on unsupported macOSes, it is needed to
switch to newer macOS to keep reasonable build times.
Besides the original check (-n $WAYLAND_DISPLAY), do not use the bundled
libraries also when there are the libwayland* in system (so not needed).
This allows `uv -d sdl:driver=KMSDRM` even from console when
WAYLAND_DISPLAY is not defined but Wayland is run in a different session.
If the WAYLAND_DISPLAY is defined but the not all libwayland* are found
in the system, now the bundled version is used (changed- was not
previously).
ilibpipewire-0.3.so.0 was added to excludelist causing the distcheck fail,
see [1]
We do not need to exclude the lib, because we already handle the preload
in data/scripts/Linux-AppImage/AppRun:282 (the case is similar as for
libjack.so.0 as also mentioned in the related AppImage issue).
[1]:
04af461f47
Use absolute path for libajantv2 library.
The relative path was used as for now, which is true in the CI (libajantv2
is extracted to .) but is less eligible for interactive use.
Built libajantv2 has now a different name:
```
2024-07-25T10:26:49.7807780Z Creating library D:/a/UltraGrid/UltraGrid/libajantv2/build/ajantv2/Release/ajantv2_vs143_MT.lib and object D:/a/UltraGrid/UltraGrid/libajantv2/build/ajantv2/Release/ajantv2_vs143_MT.exp
2024-07-25T10:26:50.0519395Z ajantv2.vcxproj -> D:\a\UltraGrid\UltraGrid\libajantv2\build\ajantv2\Release\ajantv2_vs143_MT.dll
```
Doesn't seem to be required and nwo (since the new PPA use) causes following error:
```
ultragrid_vidcap_rtsp.so: /lib/x86_64-linux-gnu/libgnutls.so.30: undefined symbol: nettle_siv_gcm_aes256_decrypt_message, version NETTLE_8
``
(obviously) when libgnutls30 is not installed, eg. in Ubuntu 24.04.
Check if bundled libavcodec has all dependencies satisfied from system
libva, if not do not use the system ones.
Fixes the error message:
```
undefined symbol: vaMapBuffer2 (./squashfs-root/usr/lib/libavutil.so.59)
```
on (some) current Ubuntu 22.04 with current continuous AI.
If run from within the AppImage, -h shows help of the AppImage wrapper
itself, not the one of UltraGrid. So add --ug-help as an alternative for
`-o uv -h`.
First considered was a shortopt -H, but this is currently mapped by UG
for --full-help (which can be used directly), so keep the long variant
in order not to make it confusing.
Since we must now build the new libajantv2, which takes some 3 minutes,
cache the build in order not to increase the build time.
Also use libajantv2 path for the SDK instead of AJA (doesn't require
unneeded rename), which will be Windows specific.
Only do it in MSW, since the build is slowest. But it can be easily used
also for other platforms.
Use libajantv2 + switch the build system to the new lib at the same time
(it is easier and there is perhaps not a big advantage to splitthis to
2 steps).
- moved ultragrid-bugreport-collect.sh to doc (will be bundled automatically)
- moved data/MacOS-bundle-template to data/template (just a move)
- added data/template/README.md
was UltraGrid-<arch>-<version>.dmg (eg. UltraGrid-arm64-1.9.2.dmg),
which is not consistent with other platforms,
eg. UltraGrid-1.9.2-x86_64.AppImage and UltraGrid-1.9.2-win64.zip. The
one with the version just after UltraGrid seems to be more natural.
In the current AppImages, the bundled libavcodec depends just on
libva.so.2 but libavutil depends on libva.so.2, libav-drm.so.2 and
libva-x11.so.2. Since the availability of any of libva-*.so was not
detected, even if those are not present (eg. in GRML 24.02 libva-x11.so.2
is missing), the system library was used.
As a solution, check the availability of all libav.* that is bundled
libavcodec and libavutil depending on.
if it happens that parent dir is /tmp, eg.:
`ug.AppImage -t testcard:file=/tmp/img.pam`
do not whitelist /tmp, otherwise following error occurs:
`Error: invalid whitelist path /tmp`
This respawns the process instead of the script. The advantage is that
when the AppImage process is killed, UG is killed as well, eg. in this
case:
```
UltraGrid-continuous-x86_64.AppImage -t testcard & (sleep 3; kill $!)
```
Otherwise just the parent process (the shell) is killed, leaving the child
(UG) running on background, which is undesirable. See also:
<https://stackoverflow.com/questions/8533377/why-child-process-still-alive-after-parent-process-was-killed-in-linux>
For the above command, `kill -SIGTERM -$!` would be required to do
the job, but the parent shell process keeps "running" in the background
and `bg` needs to be run.
Do not bundle SDL_SOUNDFONT instead of the SoundFont now included in
sources. This is a bit contra-intuitive, since the user may have set
the SDL_SOUNDFONT for its primary reason. Using the variable was just
a workaround to find _any_ usable soundfont, which is now unneeded.
Also, the user-selected soundfont may have hundreds of MB (as usually
the general-purpose soundfonts have).
The size is only some 90 kB when compressed so it is perhaps not worth
downloading from external source and depend on it.
It could be even bundled directly, not by CI. Now it is actually the case
in Linux (create-appimage.sh) but it is also possible for other platforms.
It is not needed to split the string like this:
```
a="potentially long string on"\
"multiple lines"
```
but it is better rather to put the '\' inside the string:
```
a="potentially long string on\
multiple lines"
```
