Some more testing for #976 — windows 0.17.0 r11649 against fedora 23 0.17.0 r11669.
Was scaling up and down and up and down (no sign of repro for #1087 gdk_is_colormap error from OSX, despite using --opengl=on on Intel 4000 graphics on windows bootcamped mac-mini)... and all was well until I used control-C to disconnect.
Client-side, got:
2016-01-12 13:16:57,895 received console event CTRL_C Traceback (most recent call last): File "E:\xpra\trunk\src\xpra\sound\sink.py", line 253, in add_data KeyboardInterrupt 2016-01-12 13:16:57,903 sound output stopping
Oddest part is that I don't have an E drive mounted... bootcamped xpra is in C:\Program Files (x86)\Xpra.
r11669 already fixes that, it is scheduled for backports.
Please close if that fixes it.
The path is my build path, must be an artifact of using cx_freeze for packaging the python3 sound bits on win32, it must be "freezing" the path it found at build time.
Haven't seen the issue for some time, so I'll judge it fixed. Closing.
this ticket has been moved to: https://github.com/Xpra-org/xpra/issues/1088