Author: ael Date: To: dng Subject: Re: [DNG] apulse in experimental
On Wed, Apr 19, 2017 at 01:08:19PM +0100, KatolaZ wrote: > On Wed, Apr 19, 2017 at 12:44:32PM +0100, ael wrote:
>
> > This was on a system with no pulseaudio, only alsa. Maybe putting apulse
> > in its own folder will fix things.
> >
>
> My little understanding is that if a program is compiled with
> pulseaudio support it will first check whether the pulseaudio server
> is available, and if PA is not it should revert to alsa (if
> possible). The problem is that many utilities have started supporting
> only pulseaudio, despite ALSA is still around.
>
> This is where apulse comes in the scene: if a program requires
> pulseaudio, apulse will lure it into believing that pulseaudio is
> available, but behind the scenes it will use ALSA instead. This is
> more than enough for, e.g., skype or firefox, but might not work if
> the program uses more "advanced" PA functions, but here my scarce
> knowledge of apulse comes to an end.
I had speculated that something like that might be happening. But if
apulse gets linked even when a program is not started with the apulse
"wrapper", then we have a problem so long as it only provides
partial/inadequate support. Perhaps mpv and mplayer and so on might be
compiled with only alsa support on Devuan, but that would only be a
short term fix.
Lack of time and worry about not having a functional mpv (and other
programs) is the main reason that I have not converted to devuan on
existing machines. It will definitely be on new boxes...