if no video device is detected "Preview unavailable" is just useless
as there is no video device listed.
Change-Id: I071da1e211b16f807ed22f69186a567db054c373
GitLab: #1076
Ubuntu 23.04 is not yet functional due to python2 missing.
We're working on fixing Jami for Qt 6.4 (they changed the whole video
pipeline and introduced some breaking changes).
Fedora 38 is working fine.
GitLab: #1086
Change-Id: I51bbecdfbcf908e2bf57c4c002ef4216a34d746d
The commit was reverted in 2b7a421e due to breaking some of the
packaging builds. This has now been rectified.
Change-Id: I3ea67962cc65f243f106fe00316634d580f51a52
- add an outline on the item that have the active focus
- override only after a tab key press
GitLab: #1082
Change-Id: Ib549052e0fa2fe6d781df6980ce662919925c473
The usage (--help) says: "Build with debug support; run in GDB", but
GDB was only used when using the D-Bus daemon. Run a libwrap built
Jami with GDB as well.
* build.py (run_run): Disable SIGINT when using GDB, and invoke the
Jami client with GDB. Disable output redirection to logs when using GDB.
Change-Id: Icf3415a3d1fbb87f193d0ede07cb1e1fbb179ce2
This partially reverts commit 6e0a30024bd5d0f24f8c16960a56b02737bf0131
from the former jami-project repository ("build: Improvements to the
generation of the release tarballs."), which caused the source tarball
to always be rebuilt, even in the absence of changes.
This gets in the way when testing the various packaging targets, so
revert it.
* extras/packaging/gnu-linux/Makefile (release-tarball): Do not always
clear the tarballs.manifest file. The purge-release-tarballs target
can be used for that.
Change-Id: I30ebda8dc258d0c95c6cf7e02a814627ac5b812b
This is a big time saver in the context of CI testing where we only need to supply some env vars instead.
It would also be nice to also have the env vars filled out for CMake-generated project runtime configs to prevent unnecessary deployment for development builds, but that's not done in this commit.
Change-Id: I338827cb87c9fc71c9a6b4b3076b22aa7cf7cfa3
Removes some intelligent grouping in the code. Perhaps normalization can be selectively done to the the strings and theme files at some point.
Gitlab: #1059
Change-Id: If9b0363567bc715dca0bdb4e3662612a5b159361
+ Requires that the Qt path is supplied when calling build.py --init.
+ format.py now also takes a --qt argument to specify the Qt path, used to locate qmlformat.
+ format.py takes a --type argument to specify the type of files to format (qml, cpp, or both).
qmlformat is called with `--normalize` and `--force`.
Gitlab: #1059
Change-Id: Id9ff4b17018370696792b44f55ed2f4bc8091193
This homogenizes the script calling convention with other platforms, and will allow for a cleaner approach when searching for Qt tools (qmlformat).
Gitlab: #1059
Change-Id: I540ff61bd92186bb5a2998670665631e352852b9
Devs will need to remove the existing pre-commit hook or regenerate the hook with `build.py --init`.
Gitlab: #1059
Change-Id: I1cec9150c7781d769cb229dfd170a9dcfc819edb
Also, force the format of the vcard to always be the same to
avoid to change the shasum in a useless way.
Finally, only change if there is new details
https://git.jami.net/savoirfairelinux/jami-project/-/issues/1558
Change-Id: I95bc52e13c8d1020eae6ad64e77d4e58b1ee734b
+ set the maximum number of characters in the input
+ elide the text when it's too long
+ save on focus change in the settings view
GitLab: #1064
Change-Id: I973ff09cb3ce56cda98a0b40cedde88ac3829c1d
Now that the settings are not loaded on start, the updatemanager signals need to be moved to the main window component.
Gitlab: #1056
Change-Id: I17e750ef1417e8bfca7d06976f14a1d75492cec5
