otdewiljes

Well, it's great to see that this issue is still on the radar. For my use case, it would be rendered moot if ever something GPU-accelerated became available as backend/device...