In case of conflict in i18n, keep the version of the branch "ours"
To resolve a particular conflict, "git checkout --ours path/to/file.desktop"

Summary: Use O_CLOEXEC flag.

Reviewers: #kde_applications, aacid

Reviewed By: #kde_applications, aacid

Differential Revision: https://phabricator.kde.org/D22123

Summary:
Fast blanked DVD-RW can only be burned in DAO mode. Some DVD drives
can't detect whether DVD-RW is fast or fully blanked, and the only
way to know it is from the error message from growisofs.

Currently k3b shows very uninformative message:
"Fatal error during recording: Input/output error".
This is because the output of growisofs is not parsed correctly.
First of all, stderr output is lost when growisofs is finished,
because K3b::ActivePipe closes the corresponding K3b::Process.
This is fixed by separating stderr and stdout channels of
growisofs. Secondly, the existing code assumes that the message
starts with ":-[", while it actually starts with ":-(".

Probably the bug was introduced 10 years ago by the commit
3f6bf50c.

Reviewers: #kde_applications, aacid

Reviewed By: #kde_applications, aacid

Subscribers: aacid

Differential Revision: https://phabricator.kde.org/D21880

In case of conflict in i18n, keep the version of the branch "ours"
To resolve a particular conflict, "git checkout --ours path/to/file.desktop"

In case of conflict in i18n, keep the version of the branch "ours"
To resolve a particular conflict, "git checkout --ours path/to/file.desktop"

Summary: Fix detection of DVD-RW.

Reviewers: #kde_applications, aacid

Reviewed By: #kde_applications, aacid

Subscribers: aacid

Differential Revision: https://phabricator.kde.org/D21751

Summary: QSignalMapper is obsolete and the code with lambda is shorter and clearer.

Reviewers: lesliezhai, aacid

Reviewed By: aacid

Differential Revision: https://phabricator.kde.org/D15288

In case of conflict in i18n, keep the version of the branch "ours"
To resolve a particular conflict, "git checkout --ours path/to/file.desktop"

In case of conflict in i18n, keep the version of the branch "ours"
To resolve a particular conflict, "git checkout --ours path/to/file.desktop"

Summary:
With the newer ECM requirement introduced in [1]
CMAKE_RUNTIME_OUTPUT_DIRECTORY gets set [2], which causes the tests
to fail because they don't find the test executables. Adding the
NAME and COMMAND keywords, which aren't really optional according to
cmake's docs, makes them pass again.

[1] dc40d092
[2] https://phabricator.kde.org/D7198

Differential Revision: https://phabricator.kde.org/D20662

In case of conflict in i18n, keep the version of the branch "ours"
To resolve a particular conflict, "git checkout --ours path/to/file.desktop"

In case of conflict in i18n, keep the version of the branch "ours"
To resolve a particular conflict, "git checkout --ours path/to/file.desktop"

See also: https://cgit.kde.org/k3b.git/commit/?id=026a29ea79

BUG: 405994

Summary: Now that the helper actually works it makes no sense to have it hidden behind an option

Test Plan: it works

Reviewers: dfaure

Reviewed By: dfaure

Subscribers: security-team, lesliezhai, kossebau, arojas

Differential Revision: https://phabricator.kde.org/D20135

statusChanged doesn't really seem to work, and even if it worked, it
only means the job has been authorized, not that the job has finished,
result is what you want for job finished

and not KAuth::ActionReply, since this is done via invokeMethod magic
and that does string matching

delete nullptr is fine

moc doesn't know about defines so ends up not generating the slots
even when the define is on, so just define the slot always and use
the guard only in the cpp code