After a few posts about why we release KDE 4.0 while it's in a not-yet-perfect state, Aaron wrote an excellent post about it. Even though is post was pretty much perfect, I want to add a little thing to it - maybe obvious to most, but probably insightful to others.
As many of you might know, KDE is a large project. Under the KDE umbrella are projects like Kalzium, KTouch and KStars from KDE-EDU (each saw a lot of work). Of course, there are the games, receiving new graphics and features. KDEGraphics, which contains Gwenview (video), Okular, and kolourpaint. And let's not forget the many KDE Base applications like Kate or Juk from the KDE multimedia area. And of course I'm skipping over many other useful KDE applications.
These applications are, for most part, ready for release. Actually, the games and educational applications could have had a release a year ago. The latest OpenSuse, 10.3, already ships some KDE games and KDE Edu applications from KDE 4!
We could let these applications wait for another 6 months - sure. But that would hurt them. They would lose users (no release = no new features & fixes so users will start to look somewhere else). They would not gain new developers (which are, after all, often attracted to cool projects - and you're not that cool if you don't release). They might even lose developers (developers often develop because they want their code to be used. Not sit bit-rotting in a big repository somewhere).
So, a choice had to be made. Wait more, let the parts of KDE 4.0 which aren't up to 3.5 standards mature, while other parts of KDE would slowly start to deteriorate? Or do a release in true FOSS spirit, bring it out while it's fresh, and hope that it will infuse a stream of new developers to help us make it more mature?
Well, you all know what happened. Yesterday tagging. Release in a little over a week. I think the right decision was made.