When ug process doesn't respond to termination GUI needs to kill it.
This makes the return code equal to 9 and GUI complains about a crash.
Without a message in the log this could have been confusing.
Probably all of the useful devices are now reported by capabilities, so
it's likely to be not even needed, so to avoid confusion it's moved into
the settings window and given a more descriptive name
This was a dirty fix for #106. Video display devices were reported with
the module name and device option (e.g. "decklink:device=0"). The
display devices are now reported as separate module name and device
option strings, which means that we can test for exact match.
This reverts commit dc9ede91f1.
All device configuration is now reported from capabilities in the same
way. Therefore it is now possible to generate the .device settings keys
during runtime for all reported modules
The QProcess::start() method is replaced with a new one that requires
executable path and arguments passed separately.
Each argument must be now passed individualy using QStringList and
have their spaces escaped automatically and passing argument
list as a single string would no longer work. This is now handled by
splitting the argument string on spaces.
The executable path can no longer contain quotes as spaces are escaped
by QT automatically now.
Since QTextEdit::setTabStopDistance() is available starting 5.10 and we
still support Qt 5.5 (Ubuntu 16), it's moved to log_window.cpp and
ifdefed according to version
This partially reverts f8d9ac9d. Device configuration is now passed as
an opaques string rather than to be decorated by GUI (prepended
"device=" etc.).
NDI, SPOUT and Syphon are able to capture/playback audio.
Decklink audio playback on the other hand is not dependant upon video
playback selection (works as a standalone device).
