Some KDE projects like Amarok have already switched to Git, and, according to Seigo, "They're moving at an amazing pace, and the main reason is that the number of contributors have gone up." Seigo does expect "a week where everyone's going, 'What the hell am I doing?'" because KDE has over five millions lines of current code, and some 58 gigabytes of archives to transfer. But "hopefully, when we all move over, KDE development will go even faster."
But, whatever changes are in store for KDE, Seigo suggests that they should occur with a minimum of problems. In particular, Seigo says that, because the KDE 4 series is designed to be more easily altered than previous release series, code bloat is less likely to be a problem, because fewer kludges are needed to work around hard-coded limitations.
Moreover, "in cases where there is bloat, we disable it at run time if your battery starts to run out or if your machine is too slow. You give up some of the bling, but you still have a fully functional, good system."
Seigo does express some concern about another user revolt like the one that happened with KDE 4.0. However, he blames the revolt partly on distribution's efforts to be the first with new software, and partly on inevitability.
"We have a very bad habit in free software in general that, whenever necessary development happens -- something that has to be done -- we're not good at creating new products around that," Seigo says. "We just say, 'Here's a new thing,' and we throw it out and see how it goes. The distributions really need to get better with how they deal with such things, because upstream can't realy stop and ossify our code. We need to do this every so many years, and the distributions have to learn to deal with it. And it's not exclusively a KDE thing; we saw the same thing with PulseAudio"
And when will the KDE 5 series roll out? Seigo says that the version number might change to keep it in sync with that of the Qt toolkit, but that, otherwise, a new development series is unlikely.
"We're maybe halfway through what we want to achieve with KDE 4," Seigo says. "The frameworks that we have are not being fully utilized, and wwe're not looking at things and saying, 'Dang, I wish there was something we could do.' We were really doing that with KDE 3, and we don't have any of that on our plate right now. We're kind of like a kid whose mother buys a shirt three sizes too big and tells him that he'll grow into it -- we still have lots of room for growth."
In a few years, KDE 5 will probably happen, but Seigo predicts that the change would be more like the one between the KDE 2 and 3 series, and not nearly as radical as that between KDE 3 and 4. "I think we've probably got a good decade of development in the framework we have right now," Seigo says.