Re: NLD10 and GNOME

[Kjartan Maraas <kmaraas broadpark no>]

ons, 08,.02.2006 kl. 10.55 -0500, skrev Dan Winship:
> Alan Cox wrote:
> > So if Fedora, Ubuntu and every other Gnome using distribution also start
> > doing tons of private development
> 
> (Excluding Xgl, there was hardly "tons" of private development.)
> 
> > then trying to jam it all in CVS
> > afterwards how do you expect Gnome to develop when all these variants
> > suddenely try and get merged and all overlap and clash.
> 
> We don't. A lot of people have assumed that we're expecting to force the
> new menu code into the GNOME mainline at some point, which I guess is a
> reasonable assumption given what happened with Ximian Desktop, etc, but
> that was never the plan here. At the moment we're not even planning to
> ship it in SUSE 10.1 (which is 90% the same codebase as NLD10). The new
> menu is something we did for NLD, and if the community wants it too,
> then great, but we didn't do it with the expectation that they
> necessarily would. It's like Industrial was.
> 
In a sense, but a theme is more self-contained and wouldn't need review
in full the samme way that an extension to the panel menu code would.

> > Nor does the committee argument stand up. It is perfectly possible to
> > post in advance that "we are going to do this, we've created a temporary
> > alternate repository for the work and if you want to join in or help
> > merge stuff back as it meets acceptability please sign up"
> 
> Yes, I shouldn't have suggested that secrecy was a necessary part of the
> mix. The secrecy doesn't necessarily help. But how does it actually
> *hurt*? Yes, there are integration issues in some cases, but not in this
> case. Yes, there are code review issues as you mentioned in another
> message, but it's not like the GNOME community and/or Red Hat is
> reviewing the work we do on YaST or iFolder or any of dozens of other
> non-GNOME things, so that argument also feels weak. Novell has also been
> doing tons of GNOME work in the open, so it's not like we're trying to
> get a free ride off GNOME. So what exactly have we done wrong?
> 
I don't think you've done anything wrong. Nothing that isn't weighed up
by the great contribution this is to making linux on the desktop
succeed. It just happens to stomp on a sore foot, so to speak. This is a
community problem and it's the community that has to solve it if we want
to avoid this kind of thing happening in the future. Good thing Novell
is part of the community too :-)

Looking forward to seeing some of this incredibly cool technology pop up
on my desktop too one of these days. I think also that we sometimes put
too much emphasis on never duplicating code or effort. I really think
that it gives the community as a whole more experience in how to
approach a certain problem and that "both sides" can learn from each
other's mistakes when this happens. I'm sure there are lessons to be
learned from metacity/libcm/compositor vs. Xgl/compiz that will benefit
both projects long term. There are probably other examples of the same.

I do mostly agree that you could have achieved the same step forward
codewise without the secrecy, but it would have created a fuss and you
would have lost the fun of announcing something entirely new to the
world :)

Cheers and thanks to everyone involved for doing all the work this must
have been.

Kjartan