One Gnome (was: Time to heat up the new module discussion)

[Shaun McCance <shaunm gnome org>]

On Thu, 2006-07-13 at 16:50 +0100, Iain * wrote:
> On 7/12/06, Darren Kenny <Darren Kenny sun com> wrote:
> > Just think about what happens when a user logs into a desktop that has Python
> > and C# based applets included with C based applets:
> > - The panel starts
> >   - It starts C/Bonobo based applets - the smallest of which already consumes
> >     approx 40Mb of memory.
> >   - It starts Python applets - each of these takes up approx 70Mb of memory -
> >     and very little of this is shared
> >   - It starts a C# based applet - and this pulls in Mono, which I'm sure isn't
> >     that small, but I guess at least it does share memory better than Python,
> >     but there is still quite a lot of additional memory pulled in.
> 
> Then...umm, don't run them?

A lot of us have been using this argument a lot lately, and I
don't think it's a productive way to develop software.  Sure,
it's wonderful that we have choice, and we should never take
that choice away from our users.  That does not mean that we
should force (or actively encourage) our users to choose not
to run our software.

Realize, Iain, that your reply was not just to a user concerned
about his desktop.  Darren has clearly shown in his posts that
he's concerned about how this will pan out for Sun's offerings
of Gnome.  So you're not just telling a user he's free to run
what applications he wants.  You're encouraging one of our
distributors to ship something other than stock Gnome.

This has been a brewing problem for many years now, and we need
it to stop.  We need one Gnome.  Many moons ago, the idea was
that Gnome didn't need its own window manager.  From a purely
technical standpoint, that's true, especially now that we have
initiatives like the EWMH.  To the user, though, it's another
story altogether.  Having a blessed window manager means that
we can put window management configuration in sensible places
in the control center.  It means we can document the desktop's
behavior and functionality.  It means we can actually talk to
people about our desktop without a bunch of if's and but's.

When Gnome is divided, we can't build interfaces that work the
way users think; we have to expose application boundaries that
aren't relevant to users.  We can't market Gnome effectively,
because everything we might want to say is false in at least
one of our distributions.  We can't document Gnome correctly,
because we don't know what's actually on our users' screens.

ISDs can't target a divided Gnome like they can target Mac
or Windows.  Having a solid and stable developer platform
is wonderful, but much of what makes a great user experience
comes from outside the platform.

If we don't unite Gnome, we will lose it entirely.

--
Shaun