Le Mar 18 décembre 2007 09:46, Linus Walleij a écrit :
On Tue, 18 Dec 2007, Nicolas Mailhot wrote:
> Why are we seeing the same lame arguments again and again?
Now that's a bit rude. Lennart is pushing the limits for an advanced
Linux desktop/workstation with his work, and frankly I think it's
worth it, even if it messes with some server use cases a bit.
I don't question Lennart qualifications or qualities, I question the
trend we've been seeing lately where every new Fedora tech is designed
for very narrowly defined use cases that do not build on Fedora's
general-purpose inheritance and needlessly ignores some classes of
users (which have to be put back in after a few releases).
I question the artificial separation between "desktop" and "server"
and the way some teams purposefully ignore problems of other teams and
think that's a good deal for Fedora, when in fact there is a lot of
grey area between and our strength WRT proprietary marketing-driven
competition was we were able to manage this grey area.
Which is not to say Fedora should transform in RHEL or something like
Debian that takes years to get out, but there is a huge difference
between not trying to be a perfect big iron solution and declaring
anything which has a small dhcp/bind/apache or any other sort of
service running is NOT-A-DESKTOP and thus WE-DON'T-CARE and should
There are lots of use cases even for home users (let alone SOHO users)
that involve partly serverized desktops which should IMHO be handled
well by Fedora.
I think like Linus that building something that scales from embedded
to huge server creates value for everyone involved, and going many
small specialized incompatible spins that do not play well together
and reinvent square wheels is a waste of resources.
It's a shame with our heritage when one of our teams declares stuff
which is routinely handled by non-server windows systems in the wild
have no place in Fedora's "desktop" roadmap.