Thanks @adamwilson, I had googled a bit indeed, and found some older posts about it which don’t quite translate / build with more recent versions - (I asked about it here too with no replies).
I’ve not spent much time on it yet, but yes I can imagine possible worst-case scenarios where even with the dual message loop it should be at least possible - for example the two loops talking over queues with each other, each in its own thread, or process (surely that’s overkill).
Thank you for checking in @dave96!
I know QT for plugin UI’s is not recommended but for a standalone application/host it still makes sense, so it should be worth trying to get a POC running.
I’m afraid the stack trace isn’t particularly exciting:
tracktion::engine::TransportControl::stop(bool, bool, bool) tracktion_TransportControl.cpp:904
tracktion::engine::TransportControl::timerCallback() tracktion_TransportControl.cpp:1075
juce::TimerThread::callTimers() juce_Timer.cpp:170
juce::TimerThread::CallTimersMessage::messageCallback() juce_Timer.cpp:271
juce::MessageQueue::deliverNextMessage() juce_MessageQueue_mac.h:93
juce::MessageQueue::runLoopCallback() juce_MessageQueue_mac.h:104
juce::MessageQueue::runLoopSourceCallback(void *) juce_MessageQueue_mac.h:112
main main.cpp:48
I’ll carry on experimenting.
