Re: State of gvfs in Gnome 2.21

[Alexander Larsson <alexl redhat com>]

On Tue, 2008-02-12 at 16:13 +0200, Lucas Rocha wrote:

> > But agreed that the right thing to do is to delay the release rather
> > than release a .0 with substantial regressions (as I ranted on a bit
> > at my blog and on gnome-bugsquad.)
> 
> I agree. We shouldn'd discard the possibility of either postponing the
> gvfs-based Nautilus or delaying the .0 release if needed. Obviously,
> releasing Nautilus with too many or some big regressions is not a good
> plan.
> 
> Personally, I'm more in favor of postponing the gvfs-based Nautilus
> because delaying our release can bring more problems for us and for
> the projects relying on our schedule.
> 
> Hence, it would be good:
> - to have a plan with the list of regressions we can't accept for 2.22
> - to hear (mainly) from Alex if this plan is feasible

>From my side, postponing the gvfs-based nautilus means shipping the 2.20
eel and nautilus for Gnome 2.22. This is clearly possible and is more or
less zero work for me. It is however a problem for the other parts of
the desktop that has ported to gio. I can't really talk for those
modules, but I believe it is doable to revert the gio specific changes,
although it will take some time.

I can't say I'm happy about something like that though, since I spent
the last 1.5 years or so trying to make this happen for 2.22. And if it
gets postponed the code will get little or no testing and little use by
apps, so I'm not sure it will be in a much better shape once 2.24 comes
around.

I don't personally think the regressions are that bad, and there are
some good new features and improvements that balance things. I will
continue to work on fixing bugs and regression until the release date,
and I was hoping that other people would help out. The only thing that I
think might be hard to get for 2.22.0 (especially with more help) is ftp
support, as ftp is kinda tricky to implement as a virtual filesystem
(the protocol sucks) and its not really been started.