These whole Ubuntu-driven set timed releases are in that respect a plague and lead to disasters such as KDE 4.0 and the Intel drivers in Jaunty. I see it less acutely with a distro (since it's usually just a collection of code and less its development), but a project doing that sort of thing is an evidence of immense stupidity. One shouldn't really trust KDE after that sort of thing.
I don't see anywhere written '*.0 are testing, unstable releases'. A release is the moment you should tag your development branch stable enough to be used by the majority of people + however long you need to squash normal-high bugs. Even Fedora, as we just noticed this week, takes matters more seriously than KDE. It's ridiculous.
Just follow GNOME's efficient and slow phasing out (look at the roadmap) and moving to 3.0. This is how you should move, even if it takes longer.
Last edited by susikala; 05-22-2009 at 05:52 AM.
I don't find the KDE devs at fault for picking a differing versioning style from what people are used to. After all, the only place who cares about 1.0 is the proprietary world, and even they because it's "just always been that way".
A lot of free software projects don't go with the "usual" versioning style. And it doesn't matter if one project's 1.0 is not what another project would tag with the same number.
I read the release announcement, and decided not to try it at that time. It took less than a minute, and it's not too much to ask to read a short notice is it.
I can dispute a lot about this 'efficiency', but I won't in this thread. Moving from QT3 to QT4 is like moving from Gtk* to QT4 (ok, maybe in some, but probably in big part), so in my opinion Gnomes way is not the best in this case.Just follow GNOME's efficient and slow phasing out (look at the roadmap) and moving to 3.0. This is how you should move, even if it takes longer.
Last edited by kraftman; 05-22-2009 at 10:23 AM.
What does GTK vs Qt have to do with Wayland? Oh, right. Nothing.
I look forward to seeing Wayland backends for both toolkits, not to mention Firefox and OOo and Cairo. I look forward to seeing if Wayland is actually usable as the actual desktop display server and not just as an X multiplexer, or to see if X just gets those same features and obviates Wayland. For all the complaints against X, it has kept up just fine so far, especially since the X.org rejuvenation. Time will tell, I suppose.
If there will be only one toolkit Wayland dev/devs will have less work to do The same about graphic cards drivers developers.What does GTK vs Qt have to do with Wayland? Oh, right. Nothing.
In an ideal world there'd be a stable API that all driver writers (both opensource and proprietary) can write to to ensure their drivers will perform. Sadly that is not the case now...API keeps changing and causing all kinds of headaches for graphics driver writers...
Desktop environments/window managers should also have a stable API too