[pulseaudio-discuss] Default sink and PULSE_SINK

[Михаил Новоселов]

Hello,

Here is a question from the developer of Chromium browser, which was 
asked by him after I filed a bug for regression in Chromium 69: 
https://gitlab.freedesktop.org/pulseaudio/pulseaudio/issues/573

As nobody answered on GitLab, I am quoting it here. The original author 
is CCed. Thanks in advance for answering.

------------------------------------------------

https://crbug.com/880180 was recently filed because a recent change in 
Chromium made it ignore the PULSE_SINK variable when choosing the 
default output device.
After the change, Chromium uses as default the sink obtained by calling 
pa_context_get_server_info 
((https://freedesktop.org/software/pulseaudio/doxygen/introspect_8h.html#a6f3bdd9982d9dec70ca5acf6ada3a9a2), 
which returns a
pa_server_info struct 
(https://freedesktop.org/software/pulseaudio/doxygen/structpa__server__info.html) 
containing a default_sink_name field (among others).
After this change, things work exactly as described in 
https://www.freedesktop.org/wiki/Software/PulseAudio/Documentation/User/DefaultDevice/, 
which does not mention $PULSE_SINK.
However, this appears to be in conflict with what is stated 
https://www.freedesktop.org/wiki/Software/PulseAudio/FAQ/#index5h3, 
where it says that $PULSE_SINK is the default sink.
Prior to the change in Chromium, the default device was specified using 
either the empty string or the "default" string, and the sink in 
$PULSE_SINK was used if specified. This prompted a user to file 
https://crbug.com/880180, since it broke their use case.
I would like to know if this is a bug or if there is a way to get the 
default device using PulseAudio APIs such that the PULSE_SINK variable 
is taken into account.

-- 
------
С уважением,
Михаил Новоселов | mikhailnov at dumalogiya.ru | https://nixtux.ru