With -d5 I get this in the log: > 2015-12-04T15:18:58: pulseaudio[E]: D: [tunnel-sink] context.c: Protocol version: remote 30, local 30 > 2015-12-04T15:18:58: pulseaudio[E]: D: [tunnel-sink] context.c: Negotiated SHM: yes > 2015-12-04T15:18:58: pulseaudio[E]: D: [tunnel-sink] module-tunnel-sink-new.c: Context failed: Protocol error. No idea why yet or exactly which combinations that fail to work.
The 4.4.0 client with the older PulseAudio works: > 2015-12-04T15:42:23: pulseaudio[E]: D: [pulseaudio] protocol-native.c: Protocol version: remote 30, local 28 > 2015-12-04T15:42:23: pulseaudio[E]: D: [pulseaudio] protocol-native.c: SHM possible: no > 2015-12-04T15:42:23: pulseaudio[E]: D: [pulseaudio] protocol-native.c: Negotiated SHM: no
The problem is the srb channel which is completely non-functional in our PulseAudio. Unfortunately it still tries to use it if the local PulseAudio supports it, but then of course fails. This affects any distribution which has this feature turned on, which is PulseAudio 7.0 or newer, or 6.0 with extra configuration.
Upstream bug: https://bugs.freedesktop.org/show_bug.cgi?id=93285 It doesn't affect the PulseAudio server in the sessions as it already has correct handling of failing to start the SRB channel.
Added bug 5746 about properly supporting the SRB channel.
Verified that i had the problem using old client build on fedora 23. Upgraded to client build 4969 and all works as expected.
Need to test regression using old pulseaudio on the client side.
(In reply to comment #8) > Need to test regression using old pulseaudio on the client side. Tested using thinlinc build 4969 on eLux RP (Pulseaudio V6.0). Sound works as expected.
