Use rather branch release instead of main because in main things may
break sometimes.
Eg. currently (libajantv2 commit 9060cc4a), ntv2m31enums.h is not
installed but ntv2utils.h includes that.
Seems like Wayland driver seem to need to have this set explicitly
for the eventual borders to be cleared correctly with SDL_RenderClear.
Otherwise the eventual margins may not be cleared by SDL_RenderClear
(tested with U25.10 live DVD with llvmpipe software renderer inside VM).
It seems like libbrotlidec.1.dylib is dependency of Qt but
incorrectly deployed by macdepoloyqt - just the library (that
is symlink to libbrotlidec.1.2.dylib) but not its dependency
@rpath/libbrotlicommon.1.dylib (+ fix path).
Since libbrotlidec.1.2.dylib is already deployed and fixed as a dependency
of uv (AFAIK through lavc->libjxl), just symlink that.
fixes <https://github.com/CESNET/UltraGrid/issues/478>
Currently Ubuntu 25.10 live DVD (ubuntu-25.10-desktop-amd64.iso)
does not include the library, preventing the GUI from being
run. On the other hand, if used unconditionally, it causes
the warning (even in the U25.10) for which it was removed:
<https://github.com/linuxdeploy/linuxdeploy/issues/152>
So preload the library if is in the system and use the bundled just as
a fallback.
The REST API returned 4xx on unknown object but valid JSON body. As
written originally, valid JSON was assumed (result queried on null).
So revert back the original curl calls where acquiring JSON.
- '|| true' no longer needded
- CFLAGS also doesn't seem to be required
- remove --disable-shared - when possible, linkling dynamic library may
be rather preferred than the contrary
Co not continue if SDK_NONFREE_PATH is set but deltacast
SDK not present there. At this time, this will fail later in
.github/scripts/install-common-deps.sh, anyways. Also the DELTACAST
SDK remains the last non-free SDK so it doesn't make sense to define
SDK_NONFREE_PATH GH variable while not having that SDK.
GPUJPEG commit 06f18868b message:
````
gpujpeg_reader spiff: directory read fix+updates
- FIXED: skipping content the dir entries other than EOD (+ early returns)
- pass the complete length, not -2 (it is misleading)
- length check - increase from 6 (4+2) to 8 - T.84 says ELEN should be at
least 8 (seems that EDATA should be at least 2 bytes)
```
Improves the commint 4682cf65 (2024-08-19) allowing also the capture
filter to be specified both before and after the vcap (if only one
used), eg. `-F flip -t testcard -F mirror`.
Set the bitrate only if >0. Do not set if neither set explicitly nor by
the codec preset - eg. for APV avg_bpp is intentionally set to 0. The
libopenapv uses CQP by default and setting the bitrate explicitly
disables that.
Note that this setting doesn't do anything important - setting the
AVCodecContext::bitrate to 0 does nothing. This change is avoid displaying
"Setting bitrate to 0.0 bps." message.
setparam callback is free to alter the copy, which is then used for the
actual setting.
blacklist_opts is therefore no longer needed - the formerly blacklisted
option can now be erased from the lavc_opts copy directly.
not all compat/ headers are marked as exporting, 2 reasons:
- some compat define own symbols, like platform_semafore or aligned_malloc
(not _alloc as in C11)
- some headers are too generic to export all symbols, eg. unistd.h
or windows.h
The includer needs no longer use IWYU pragmas for strcasecmp in *NIX.
In Windows, _stricmp is defined in string.h, which is not currently
exported (it contains more symbols so it isn't specific enough, in
contrary to POSIX strings.h which contains just strcasecmp and ffs
family).
The original sample code uses cascadding if-ladder evaluating Result. The
derived code was just slightly derived but it was maybe even worse because
the error handling (message) was coupled in one if/else with subsequent
API call.
This is mostly a refactor but:
1. two-times VHD_DV_SP_REFRESH_RATE removed
2. perhaps fixed misaligned error message prints to wrong calls, eg. when
board type was not HDMI, the error was print even though the previous
call didn't fail (but continued, anyways)
- delta_board_type_is_dv takes the VHD_BOARD_TYPE now
- to acquire the type from board index, delta_get_board_type() is now
externally linked
- delta_board_type_is_dv: handle types with switch (compiler warns about
unhandled enum values)
- changed compat macros a bit (works at least for VHD 5.19, 6.{19,20,21,30,32})
HDMI/DP devices are listed in the card list (and vice versa - SDI in
deltacast-dv) but are not handled so make at least a notice about that.
Doesn't seem to be necessary for deltacast-dv (from the namw the
specialization is more clear).
+ removed excess NL after card listings
