Xfce 4.12 Planned For March, GTK3 Still Uncertain

[Rallos Zek]

I'd say the really, really frightening thing is the situation with GTK. That was already apparent at last year's Desktop Summit, where Benjamin Otte described his team as mostly just him, told the audience that GTK3 had some very bad issues (with styling, iirc) that could only be helped by doing a GTK4 really soon -- and that Windows and OSX support were just not coming along.

That sounds like bad news for a couple of important and widespread projects, like inkscape, gimp or mypaint. And I couldn't care less about gnome, the project or the desktop, but those applications are important to so many people and have so much effort invested in them, starving their foundations is scary.

The XFCE devs are smart sticking with GTK+2. GTK3 is dead in the water.

[zxy_thf]

I'd say the really, really frightening thing is the situation with GTK. That was already apparent at last year's Desktop Summit, where Benjamin Otte described his team as mostly just him, told the audience that GTK3 had some very bad issues (with styling, iirc) that could only be helped by doing a GTK4 really soon -- and that Windows and OSX support were just not coming along.

That sounds like bad news for a couple of important and widespread projects, like inkscape, gimp or mypaint. And I couldn't care less about gnome, the project or the desktop, but those applications are important to so many people and have so much effort invested in them, starving their foundations is scary.

The XFCE devs are smart sticking with GTK+2. GTK3 is dead in the water.

Are you serious?

Sticking with GTK2 = No Wayland in XFCE. Unless they are trying to develop the Wayland backend for GTK2.

[LightBit]

They showed the exact numbers.
http://mail.xfce.org/pipermail/xfce4...ay/029784.html

That's a lot.

[bwat47]

They showed the exact numbers.
http://mail.xfce.org/pipermail/xfce4...ay/029784.html

Those numbers don't even necessarily mean worse performance. Slightly higher memory usage != slower performance unless you have a particularly low memory machine.

[gururise]

They really need to start supporting the newer technologies (GTK3, Gnome 3 Panel + Unity Indicators), otherwise XFCE will fall to the wayside as an alternative desktop. Whether they like it or not, most new development will occur on Gtk3, including Wayland. By sticking with an older, undeveloped toolkit, they are only hurting themselves and their users.

[gururise]

Screw them, I already ported over half their core desktop apps to GTK3 over a year ago. It took barely two evenings of coding. The git branches are still available, and they are well aware of them. They're just sticking their heads in the sand, not being cautious engineers.

If it took you just a few evenings of coding to get half their core over to Gtk3, what is the cause of the delay?? Are they just being lazy? Is there no manpower? No interest? It seems most users are interested.

[mrugiero]

I can't find any technical justification for rushing like lemmings to GTK+ 3. Few app's will benefit.
The LTS versions of the most popular distro's will be required to support GTK+ 2 for years to come.

LXDE dev's are not in a hurry : http://comments.gmane.org/gmane.comp...de.pcmanfm/239
XFCE dev's are not in a hurry : http://mail.xfce.org/pipermail/xfce4...ay/029843.html
Chromium still on GTK+ 2 for a long time : http://code.google.com/p/chromium/is...etail?id=79722

So apart from faddism I'm at a loss to justify this heavy breathing & sweating about porting like crazy to a less mature, less memory efficient toolkit.
Olease explain?

Well, there are panel plugins. XFCE hasn't a lot of good plugins itself, most of the good ones are from gnome. Since gnome 2 series is already dropped, I'm pretty sure it's plugins are unmaintained.
I'm in a dilemma about supporting migration or sticking to GTK2, since I want Wayland and updated plugins, but I'm all for performance and memory efficiency too, and GTK3 doesn't seem to care about that a lot.

Those numbers don't even necessarily mean worse performance. Slightly higher memory usage != slower performance unless you have a particularly low memory machine.

IIRC, XFCE is targeted to low resources computers. That means switching to a more resource hungry lib is usually a bad choice.

[GreatEmerald]

It would be interesting if they just ported everything to Qt5 instead Although it's highly unlikely.

Well, they are porting things to GTK3, but at a slower pace. No need to rush, really, it's better to get everything done correctly instead of just porting as soon as possible.

[stqn]

Sticking with Gtk2 is the correct decision. I thought Gtk3 had technical reasons to be made, but it turns out it’s just much worse than Gtk2 on every front… 50% more memory usage is not “slightly more”… Also it’s pretty hard to find any dark Gtk3 theme at this point.

[chris2kari]

They really need to start supporting the newer technologies (GTK3, Gnome 3 Panel + Unity Indicators), otherwise XFCE will fall to the wayside as an alternative desktop.

No, they don't actually...
There is no pressing need to urgently dump mature, stable, memory efficent design and run with our pants down to the, apparently, one man project of GTK+ 3. The dev himself is saying it should be scrapped and work started on GTK + 4 because its so bad..

Whether they like it or not, most new development will occur on Gtk3, including Wayland. By sticking with an older, undeveloped toolkit, they are only hurting themselves and their users.

Again, Wayland is an emerging long term replacement for the X Window system and as such will someday be very nice when its finished... but there is no pressing need to run to this right this very minute for the majority of desktop users.

I suspect people pushing this "rush rush" line are the diminishing breed of hard core desktop gamers or people with version number envy.