Last month's release of the KDE 4.4 desktop environment restored the faith of many after what has been something of a roller coaster two years. Even the most loyal KDEians found that loyalty stretched by the debut of version 4.0 in 2008. Reassessing, and with hindsight, I think it's fair to characterize the whole saga as both a cautionary tale and an admirable example. Admirable, because the KDE team has displayed a quality that is so often lacking in open source software development: leadership.

By 2008, thanks to its stability, performance and feature-set, KDE3 had become one of the best desktops on the computer scene. However, it was becoming clear that a successor, KDE4, would represent a significant break with the past - the best was about to get even better. More than a mere upgrade, it was to be a complete rewrite and absolutely packed with mouth-watering technology. To the credit of the development team, they released largely on time, but it was a release that pushed the community first to bafflement and then resentment. It was unusable - and when I say unusable, I mean it literally couldn't be used for any serious work - due to stability and performance problems. To most people, a 4.0 release means a new version of a finished product, yet the general consensus was that KDE4 ran like an alpha. To top it off, it lagged behind KDE3 in terms of features.

The other jarring aspect of KDE4 was that the changes that had been made were so bold. For example, music player Amarok had been an application that had evoked the sacred phrase, "killer app", magic words that desktop Linux desperately needs to have associated with it. It was, however, like so many of the old favorites, nominated for a rewrite and redesign. It's much improved since the initial, disappointing release, but it still doesn't quite have feature parity with the final KDE3 version. Konqueror, the file manager/web browser combination was another beloved, if slightly eccentric, KDE fixture that found itself depreciated in favor of something new.

Just what were team KDE thinking in terms of pushing me out of my comfort zone? The answer: they were doing what had to be done. KDE had reached the end of the line and needed to be restarted to keep up, and begin to lead, in the world of desktop environments. Some of the decisions were criticized, but there's a saying that good leadership is always in dispute. The problems that did crop up were due to poor communication and a determination to release on schedule, no matter what.

Two years on, things are back on track. It's perhaps time to begin a campaign of telling people who left KDE, "it's safe to come back!". As it stands, not only is KDE 4.4 a superb desktop, thanks to the new frameworks that are now in place the potential for new developments is almost overwhelming. Personally, I wouldn't hesitate to recommend KDE4. However, it's been a rocky couple of years getting here.