[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-multimedia
Subject:    Re: aRts as KDE2.0 audio server
From:       Stefan Westerfeld <stefan () space ! twc ! de>
Date:       1999-09-05 19:34:45
[Download RAW message or body]

   Hi!

On Sun, Sep 05, 1999 at 07:53:57PM +0200, Martin Vogt wrote:
> KDE 2.0 needs something better than "kaudioserver" which
> is in KDE 1.0 right now.
>  
> But my vote goes to *esd* as a solution for KDE 2.0.
> 
> * Esd compiles without any gtk/gdk/gnome dependencies.
> * Gnome uses esd and thus is a bit tested.
> * Its easy for a programmer to use the esd interfaces.
>
> 
> Nevertheless Stefan Westerfeld is right, that esd is not
> very flexible.
> 
> KDE should have something better (than esd).
> 
> But not for KDE 2.0!

Well, using esd is fine with me. If people want that.

> If KDE 2.0 is someday out, I think a lot of CORBA porting will
> start: The authors of the different apps like kmp3,kmikmod,...
> will do CORBA interfaces to their apps.
> Letīs wait until they have a bit experience with CORBA, and start 
> using it, before we push now a CORBA based interface into KDE
> which no one (except arts) uses.

First of all: I don't want to have a CORBA interface in KDE which
no one, but arts, uses. (In fact, I now read your sentence three
times, and am still not sure if I have understood what you think
I want to do, and why this is bad?).

The idea is to have arts in KDE, and make it provide services to
applications that are useful. This should - for the start - be all
services that esd provides as well, with a similar wrapper library,
to make things easier.

On the other hand, your idea seems to be: first wait for KDE2.0,
then later make a KDE3.0 multimedia interface.

I think there is one major problem in your idea: you just can't let
everybody add some kind of CORBA interface that just does something
or other. If you do so, you end up with a houndred incompatible apps.

KOffice wouldn't work, if some day Reggie said: oh well, I will add
some kind of interface to KWord, and some other day Torben would have
done something CORBA-like for KSpread, and well, KIllustrator couldn't
talk to both.

Only because Torben did a good job at making his OpenParts framework,
KOffice now works great. Because all that applications stick to the
same rules for embedding each other for instance.

Of course, you don't get a guarantee that if you start with aRts as
audioserver, and then slowly try to make everything in kdemultimedia
integrate nicely with arts, using CORBA, you'll get reasonable CORBA
interfaces. But if it works, then you can release KDE2.0, and have a
good foundation for KDE2.0 multimedia apps.

The idea I have is

- you take working components (that is arts and kdemultimedia)
- you slowly integrate them, without breaking something major
- with the experience you gained from that, you slowly improve the
  interfaces, the internals and the concepts
- starting from audio, you also start with midi and finally do video
  as well

That seems better to me than the approach: after KDE2.0 is out, we
wait a while, ask everybody what a kde multimedia framework should
do, integrate all that into great specifications (any volounteers
on that), and create the ultimate interface that all apps shall support
in KDE 3.0. I mean, we can still do *that* if the KDE2.0 multimedia
stuff has shortcomings that can only be removed by a complete redesign.

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

[prev in list] [next in list] [prev in thread] [next in thread]