Re: KDE Interop [Was: D-BUS background]

[Havoc Pennington <hp redhat com>]

On Sat, Mar 01, 2003 at 10:55:42AM +1100, Jeff Waugh wrote:
> 
> So, from a devil's advocate viewpoint [1], why do we need to work hard on
> interoperability with KDE? How will deep interop enhance GNOME's place "in
> the market"?
> 

It grows the overall market by showing that we have a unified desktop
platform. Nobody wants to write apps that are
desktop-environment-specific; even lots of open source application
authors have been big fans of the shared system tray spec, for
example. Because they don't want to limit their userbase to 1/2 of
UNIX desktop users, they don't want to bet on whether GNOME/KDE will
"win," and in general they just want their app to work.

Lack of interoperability means extra work for everyone from GNOME/KDE
themselves, to operating system vendors, to third party app
developers.  Not to mention causing tons of user-visible problems that
are pretty much just *bugs* from an end user's point of view.  This
slows down free software on the desktop.

On the other hand, when we can point to "GNOME and KDE are both using
this," that is a killer argument for getting wider adoption by other
projects. For example, being able to say "Qt and GTK+ are moving to
fontconfig" was crucial to getting Mozilla to move to
fontconfig. GNOME/KDE sharing EWMH has been crucial for getting
toolkit and application support for that spec.

When we can say both desktops agree, that is a convincing argument
that starts things moving. We *do* need adoption by other projects:
there will always be apps that aren't pure GNOME apps, for one reason
or another. WINE, OpenOffice, Mozilla, Qt. Those things should work
*well* under GNOME, even if pure GNOME apps have a nicer UI. This
only makes GNOME stronger.

In convincing people to use stuff, good docs and a specification are
also quite helpful. Because it implies long-term stability.  The
process of working with KDE helps keep us from cutting corners on
proper docs, specs, and modularity.

In most cases, the interoperability process has in fact improved the
technical quality of our stuff. e.g. the new menu spec is
substantially better than what's in GNOME 2.2. EWMH is *far* better
than the GNOME WM hints. There are some exceptions, but I don't think
serious ones.


If the overall userbase grows, to enhance GNOME's place in the market
we simply have to be the best choice for users (including
organizations, and sysadmins). Then new users will choose GNOME, or at
least more new users will choose GNOME.

If we don't think new users will choose GNOME, we don't deserve to
have our place in the market grow anyway. It's not like we're a
for-profit business. We are trying to make useful software and that's
pretty much the only goal. So if no one wants to use it, we should all
pack it in and do something else.

I don't think we should ever forget that the goal is to make free
software a viable option for millions of desktop users. The goal isn't
to get the GNOME logo on every desktop by any means necessary - it's
to make GNOME so good that it *earns* a place on every desktop by
being the best. If that happens, everyone here deserves to be damn
proud.

Havoc