Re: patch to esound to use ARts (KDE sound subsystem) if it's running

[Stefan Westerfeld <stefan space twc de>]

   Hi!

On Sun, May 19, 2002 at 06:17:38PM +0500, Vlad Harchev wrote:
> >  I will ask the author about this, but the reasons could be of course not that
> > compatible programming models of esd and arts (or probably there is no C API
> > for arts that supports all functions needed for emulation) and (less likely)
> > desire not to increase data segment of apps using libesd signtifically (by
> > pages of libartsc and libs it depends upon).
> 
>  Just checked libarts C api - it's too simplistic (it's stream oriented, e.g.
> there is no way to save a sample on server and then play it instantly when
> it's needed).
>  So it seems the only way that didn't require to extend arts C api (and
> probably arts model and C++ API too) was to modify esd and hijack
> audio streams that were going to hardware.

You might want to look at what is missing in CSL for a complete esd emulation.
CSL supports output via aRts, but is a bit more powerful than the aRts C API.
Tim Janik and me designed CSL specifically for being able to do the following
things:

 - provide a unified, sound server independant API for applications to output
   sound streams and samples
 - be able to support several ways of outputting the sound (artsd, plain OSS,
   and whatever else people might need (asd? alsa? esd?))
 - be as complete to allow writing a faked libesd on top of that

The latest release is at http://www.arts-project.org/doc/csl-0.2.0.html, and
although its not finished, I think this might be the best reasonable long term
strategy to solve the interoperability problems that are currently caused by
different approaches to sound I/O.

   Cu... Stefan
-- 
  -* Stefan Westerfeld, stefan space twc de (PGP!), Hamburg/Germany
     KDE Developer, project infos at http://space.twc.de/~stefan/kde *-