Comment 6 for bug 1525738

AutoDJ has always shown various kinds of weird behaviour to me and I try to avoid it. Maybe the whole implementation needs a rework.

The violated assertion in mixxx.log.4 might be caused by AutoDJ, although the logs do not reveal the actual cause:

   Fatal [Main]: ASSERT failure in QList<T>::operator[]: "index out of range", file c:\mixxx\environments\2.0-x64-release\build\qt-everywhere-opensource-src-4.8.6\include\qtcore\../../src/corelib/tools/qlist.h, line 477

I see many failures of CachingReader trying to read audio data in the more recent log files mixxx.log.2/.3:

   Debug []: ERROR: Could not submit read request for ...

This might be the cause for the truncated files and the confusion of AutoDJ. The implementation of CachingReader(-Worker) gets a rework for 2.1 together with the new SoundSource backend. I've never seen any failure of CachingReader(-Worker) recently in my logs since I'm using 2.1 alpha.

Unfortunately those changes cannot be backported easily to 2.0. We need to find out the cause for the truncation and implement a quick fix for 2.0 until 2.1 is ready for prime time.