5 Jul
2022
5 Jul
'22
10:14 p.m.
On Tue Jul 5 20:34:08 2022 +0000, Claire wrote:
Though the order of calls to `pulse_probe_settings` wouldn't actually change either…
After restarting PulseAudio a bunch of times (after trying to reproduce the issue with pipewire-pulse too), it eventually ended up in a state where I got the exact same test failures. But those were reproduced on `master` too, and audio input did not work in any other application either, so while it's still possible I introduced a regression, I suspect this might just have been PulseAudio being stuck in a weird state, and might not be related to the code changes at all…