PulseAudio info needed

[stan]

Dennis J. wrote:
> On 11/27/2008 02:53 PM, Mads Kiilerich wrote:
>> The rewritten PulseAudio sound server utilizes new features from the
>> low-lovel audio drivers, and currently it reveales that there is
>> problems with some kernel / alsa drivers. PulseAudio might have trigged
>> the problems, but now it is just the messenger about problems that must
>> be fixed elsewhere.
>>
> 
> I think the problem here is that most Distros/Users still use regular 
> ALSA output which now seems to use a different codepath than what 
> PulseAudio uses. As a result the bugs are hard to file against ALSA and 
> get lower priority because the driver technically works for most users. 
> If the ALSA developers could unify this codepath so that the 
> "traditional" way of accessing the driver is basically a layer using the 
> new functionality that codepath would probably receive more attention.
> 
> Regards,
>   Dennis
> 

And my suggestion is that pulseaudio should be integrated 
into ALSA.  That is, it becomes invisible to most users, who 
think they are just using alsa.  Alsa would have calls that 
allow the underlying apis that haven't been abstracted into 
pulse to be called through pulse or to disable pulse if 
desired.  Or maybe alsa just becomes part of pulse, 
depending on your advocacy.  :-)

In other words it would be a drop in replacement for dmix, 
the current component of alsa that does part of what pulse does.

The goals mismatch between alsa and pulse disappears under 
that scenario as well, as they are all just working on 
sound, not pulse or alsa.