This is already old news, but I haven't commented here yet. We already hinted this possibility in my previous blog entry on this topic, but sometime last week, we made it official.

After speaking to some people upstream, we got the impression that the GTK situation was way too risky to do a GTK 2.10 migration, with no hints on when the file selector problems would be solved. As of today, and two GTK 2.10 releases later, not all of the issues appear to have been resolved in this branch, so we may have chosen the right path.

So, with this information in our hands, we described the whole situation to the release managers, explaining what the options were, and they, of course, had no doubt on what was better for etch.

The last two months before the release, we'll try to polish the last few bits that we'd like to improve in the current 2.14 packages. For example, Joss just made a change to the session manager, to make it possible to save the user's session easily, a feature which was present until GNOME 2.12, then removed in 2.14 with apparently no sane replacement of saving sessions available for the user.

I must admit I'm a bit disappointed about not being to ship all the work we've been doing with GNOME 2.16 in experimental, although I believe it was the right choice. If the etch release is delayed for some major reason, and let's hope it's not, that might open a window to see a transition going on, if the fixes are finally in and we consider our packages release quality. If not, we're sorry, but we won't be able to sell the “latest GNOME version” argument in our release PR. ;)

The Debian GNOME team has already been talking about doing a “semi-official” 2.16 backport for etch though, so people can use stable with the current GNOME, at least for a few months. We'll see how it goes...