Originally Posted by pootle
So unless you want the application specific volume settings pulse does, it probably makes more sense to go straight to alsa.
I typically have just one sound generating application running at a time
(did I already mention that I like to keep things simple?).
But on the other hand, I'm wondering that the individual volume setting
hasn't yet been implemented on ALSA level. I'm not an ALSA expert, but
from what I have seen so far in the code, there is nothing that would
prevent somebody to implement this.
I understand that apart from the individual volume setting, pulseaudio
acts as a networked sound server. A typical "local" user (the "L-user"
aka luser, SCNR), doesn't need this functionality, so the sound system
should be hierarchical! Means: for each local sound applications, an
ALSA channels is opened that can be volume controlled individually.
And for each remote application, the (multi-threaded) sound server also
opens individual ALSA channels that can also be volume controlled
individually. The sound server could be started on demand i.e. via xinetd.
I think this would be the much better approach!
So from that perspective, pulseaudio is the plain wrong approach, IMHO!
We are getting a little bit off-topic, here :-)