Re: First release of MAS available

[Thomas Vander Stichele <thomas urgent rug ac be>]

Hey Elliot,

> > Well the question is if asking apps to depend on gstreamer is a lesser evil 
> > than needing to maintain two sets of sounds system interfaces.
> 
> Which two interfaces are you referring to that you think would need to be
> maintained?

I get the feeling that you are arguing the case of "if there was one 
really good audio server, we would all be using it and there would only be 
One True Interface".

While that would be great, it just isn't the case, and over the years it 
has been shown that we're getting farther away from that possibility, not 
closer.

I use Red Hat, which of course is a system you know, and Red Hat, as far 
as I can make out, uses OSS by default.  While, if I understand correctly, 
you still argue for EsounD in it's current form.

I'm not convinced by your argument that promotes the "choose one and stick 
to it" if it's obvious that people like you and me, who have a pretty good 
idea about the pro's and con's, in their actual daily usage have two 
conflicting output systems out of the box already.  My Gnome theoretically 
would be using esd, but my system comes with OSS.

So, to answer your question : we're at least thinking about two different 
interfaces at the very least already, OSS and EsounD.

> Although gstreamer may be different, for Gnome, it's not important to be
> able to have configurable support for multiple sound systems, only to
> integrate well with one good one (hopefully MAS).

... so, as I've said, Gnome already has support for multiple sound 
systems.  It might not be important, but it is the case right now.

> Also, if the gstreamer solution makes sense, it makes sense whether or not
> MAS is useful.
> 
> Also, depending on gstreamer will may introduce portability issues for 
> Gnome, due to cothreads. Will it?

Try and see.  Personally, I think it's not that big of a deal.  We have a 
new scheduler being worked on, NTPL is making our life that bit easier, 
and I'm pretty sure that if only one very good thread hacker looks at it 
and tells us what we might be doing wrong (you know, someone with 
knowledge that goes slightly beyond the publically available docs), it'll 
be fixed in no time.

In the past we have asked for it, in the future distributors might be 
forced to take a look at it because their users demand stuff works :)

So whichever comes first, someone looking at gstreamer threadwise because 
they want to help us out, or someone looking at it because it's their job, 
we'll deal and be happy either way.

> I think we both agree that simpler is better, I'm just not seeing using 
> the gstreamer layer as helping simplicity at all,

No problem.  I respect your point of view.  I hold Iain's in slightly 
higher regard because he actually coded an application with it though and 
experienced the advantages.  As I've said, it'll be the app developers who 
decide in the end, not the people commenting from the side of the road.  

This isn't meant derogatory (I'm pretty tired so having trouble expressing 
my views), because sometimes I hate that as well.  But given the choice 
between my perfect desktop that I am not going to ever finish by myself, 
or this very cool Gnome desktop developed by everyone together, I know 
which one I want to be putting my money on.

Thomas

 -- 

The Dave/Dina Project : future TV today ! - http://davedina.apestaart.org/
<-*- thomas (dot) apestaart (dot) org -*->
Now, if it were called the Orgasmator, 
I'd be the first to try your basic 
button-press approach ...
<-*- thomas  (at) apestaart (dot) org -*->
URGent, the best radio on the Internet - 24/7 ! - http://urgent.rug.ac.be/