(I haven’t blogged for a lot of time, sorry about that, too many things to do)

This is a short summary of the status of Xfce (especially 4.10) in Debian and Ubuntu.


To sum up: 4.6 is in squeeze, 4.8 in testing/sid/wheezy, 4.10 core in experimental.

We packaged 4.10 in experimental, but it only covers core packages (apt-get install xfce4). Goodies that haven’t been rebuilt since 4.8 (e.g. panel plugins) will not work with 4.10. Don’t upgrade if you need a special panel plugin and don’t know how to rebuild it, or don’t know what you’re doing.

It’s very likely that the next stable release (wheezy) will stick to Xfce 4.8. The main reason: direct upgrades from 4.6 to 4.10 are unsupported, untested, and very glitchy (xfce4-panel and xfce4-session don’t like that very much, you’ll experience problems to close your session from your Xfce desktop after the upgrade).

Another important reason is the date of the freeze (expected in June). The Release team is currently fighting with transitions — a backlog of two months according to some of its members —, and to “help” them some maintainers start uncoordinated transitions (e.g. mysql, gcc-defaults).


Executive summary: precise has 4.8, quantal 4.10.

People wonder why “we chose to stick to 4.8 for the LTS, as 4.10 release was very close and known”. Actually, the release schedule of Xfce 4.10 has been uncertain for several months, and only clarified around FOSDEM. That wouldn’t have left us much time to package/test/check for regressions (hint: LTS) before the feature freeze. That’s why I chose to keep Xfce 4.8, and of course you’re free to think I was wrong.

Still, I wasn’t so lazy, and backported several useful features from xfce4-settings and xfdesktop 4.9.x to precise. Also, most of the changes included in 4.10 have been made by one single person, that’s not a big number of active core developers. The current Thunar developer is also very busy at work, that’s why the improved sidepane rewrite hasn’t made its way into 4.10.

Obviously, Xubuntu 12.10 will ship Xfce 4.10 (it’s already been in quantal for a week). I maintain a PPA for precise too, many people are using it, and I didn’t receive too many complaints. ;-)

Xfce 4.12 and future

The current plan (feel free to read the Xfce development mailing list) is to keep using gtk+2 for Xfce 4.12. Xfce developers experienced an increased memory consumption when they ported and tested some components to gtk+3. Some of them also feel it’s less fun for them to spend another development cycle only to port to gtk+3.

With my Xubuntu developer hat on (all distribution packagers will probably feel the same), I can say this will make my life a nightmare until the next LTS (and Xfce 4.14). Some Canonical/Ayatana/Unity people are very good at trashing “outdated technologies”: one insane person, with apparently no knowledge of the Ubuntu archive — he has upload rights, that’s very reassuring —, even proposed to drop gtk+2 from the archive for 14.04 (for reference, it took more than 5 years to transition away from gtk+1), of course without offering his help to attain that goal.

So, yes, I very much disagree with the decision to stick to gtk+2 for Xfce 4.12, but I’m not the one maintaining core components, nor doing the porting, and if active Xfce developers don’t find that fun, you can’t force them, unless you want them to burn out, and you know I don’t want that. :-)