Re: Metacity, Mutter, GNOME Shell, GNOME-2.28

[Owen Taylor <otaylor redhat com>]

On Mon, 2009-03-30 at 15:20 -0500, Ted Gould wrote:
> On Mon, 2009-03-30 at 20:23 +0100, Alberto Ruiz wrote:
> > 2009/3/30 Ted Gould <ted gould cx>:
> > > On Wed, 2009-03-25 at 12:07 -0400, Owen Taylor wrote:
> > >> So, basically, no I don't see a way that GNOME Shell coexists with
> > >> Compiz other than as two separate shells for the GNOME desktop.
> > >
> > > And I think that coexistence is part of the problem with GNOME Shell
> > > becoming the default GNOME interface.  Distributions need something that
> > > can gracefully decline between a composited and a non-composited
> > > environment.  Not saying that Compiz can do that today, but we
> > > effectively get that with the combination of metacity and Compiz and
> > > lots of nasty hacks.  But, overall it works.
> > >
> > > For a GNOME Shell like project to be successful it will need to have
> > > either two backends or some sort of architecture that would allow for
> > > GNOME Shell features to be integrated in other less featureful
> > > shell-like tools.
> > 
> > I don't get why that statement is true. For a GNOME Shell project to
> > be successful, it hast to be freakin good.
> > Mac OS X and Windows XP are way far more successful desktop
> > environments than GNOME or KDE are, and they don't even have the
> > notion of swappable windows managers, and if they do, none uses them.
> > 
> > So what's your point here?
> 
> Swappable Window Managers isn't important.  Being able to have graceful
> degradation down to non-composited environments is.  To be entirely
> honest, some of these are problems of our own situation.  Neither Mac
> nor Windows have to worry about shipping binary nvidia drivers either.
> I'm not a fan of the situation, but we've solved this problem in the
> past with swapping window managers.  I don't think that's the only way
> to do it, but it's definitely the easiest today.
> 
> I'm not sure of all of the demo CDs out there, but I don't think that
> almost all of them come up in non-composited environments on the vast
> majority of hardware.  Having the demo be entirely different than what
> you get when install seems like a really bad idea.

It doesn't sound like a good idea to me to have a demo CD that demos
something that isn't the same thing that you get when you install. If
someone has a nice new njvidia card, what incentive do they have to
switch to Linux if the demo CD they try out looks like something from
1998?

Nouveau is coming along nicely. As is 3D support for the latest AMD
cards. I have every expectation that in a year (which is when we are
really talking about GNOME Shell starting to be widely deployed) we will
be able to support it with open source 3D drivers on the vast majority
of what people are running. Including netbooks and other low-end
devices.

What isn't going to be able to run GNOME Shell is really old
hardware ... if you have a Rage128, GNOME Shell isn't going to happen.
But when we are talking legacy hardware of that age, I think a special
purpose environment makes sense. Whether it's XFCE, or metacity +
gnome-panel, or something else.

Of course, low-graphics fallbacks that integrated seemlessly with GNOME
Shell would be nice to have. But given the reality that they would be an
enormous amount of extra work, and would seriously compromise the design
vision, they don't make sense.

- Owen