Re: GNOME 2 Sound Architecture and APIs?

[Alan Cox <alan redhat com>]

> playing mp3s".  For this limited domain of desktop applications, 
> it works ok.  It really needs a resampling overhaul (borrow one 
> from a decent MOD player, GPL rules), and portions of the 
> implementation are total crap (passing sockets, etc.).  This is
> not meant to be an exhaustive list of problems with esd, just 
> the most pathetic ones.

I think you hit most of the critical ones there. 

> I claim that for 99.999% of the average desktop users (i.e. not
> audio power users), esound is "good enough", especially if a

Here I have to disagree

> users, *no* process that sits between a high-end audio app and
> the sound card will be acceptable.  This is why esd provides a
> way to release the audio device for programs that need finer
> control (e.g. quake).  Although the Loki guys seem to work with

A sound daemon really has two jobs to do. One is to play sounds the other
is to arbitrate resources. What I mean by that is given 4 channels of
audio there is a role to play deciding who gets channels. Thus you might
want to give a channel to quake to use raw, keep one for beeps and boings
and hand the other two out to apps as needed. That implies apps giving a lot
more info when they ask for an audio channel.

The other thing it lacks is synchronization. That problem is an important one
that won't go away


_______________________________________________
gnome-hackers mailing list
gnome-hackers gnome org
http://mail.gnome.org/mailman/listinfo/gnome-hackers