Recently at a Linux show, John Littler saw a preview of a new version of KDE running on a KDE developer's laptop. The interface looked cleaner than before, and apparently there was a whole raft of new stuff under the hood. John recently interviewed KDE developer Aaron J. Seigo about the forthcoming KDE 4 (due in the fall) and also a little about the recent controversy surrounding the porting of KDE to operating systems other than Linux.
John Littler: The new features I remember include a bucketload of bug fixes, a universal messaging client, and a big cleanup of the GUI. Are there more things that you should mention?
Aaron J. Seigo: For KDE4, there is a lot more than just this. We are:
selecting a new multimedia system (aRts is unmaintained and not providing what we need)
JL: Right. I remember Stefan changed his thinking on that approach. What are the main alternatives now?
AJS: GStreamer, NMM, and MAS. GStreamer looks like the most probable candidate at this point. We'll be:
JL: I think one of the general aims has been to make the whole thing "lighter" and faster. In achieving something like that in a reasonably large codebase, what sort of methods do you use?
AJS: There is no single silver bullet, nor even any really simple answers. There is work going on at pretty much every level, including things such as:
COMPOSITEimprovements and EXA)
Add to this the hard, thankless work of old-fashioned profiling and optimization, which we do with each release, and it's a promising recipe for success.
JL: In the early days of KDE there seemed to be a definite attempt to make Windows users feel at home. Do you think you're traveling away from that now?
AJS: Windows users certainly felt (and still feel) at home, but that's different than that being an explicit goal. We've always tried to present a reasonably coherent and integrated desktop that is well behaved, which is also what people used to Windows have come to expect. Inasmuch as that remains the goal of both KDE and Microsoft, I don't expect there to be any large departures from the phenomenon of Windows users enjoying KDE.
At the same time, we certainly are willing to try a greater number of new things, which seems to be the same with Microsoft and Apple as well. I think this may be due to desktop computing entering a new phase of its evolution and maturing further, combined with the fact that KDE has come to the point where we have a solid base and have on several fronts little place else to go in terms of improvement other than to start innovating more.
JL: "Replacing the desktop and panels with a new application called Plasma": can you give me more on this? I'm not sure what I saw at Linux World, but it looked very clean.
AJS: That's part of the idea, yes. The central concepts are workflow and beauty. Some of the ideas include:
COMPOSITEare used, when available, to provide performant and beautiful alpha blending, window thumbnails, and more
JL: Any chance of a sneak screenshot?
AJS: Not at this point, sorry.
JL: What improvements result from SVG use?
AJS: Resolution-independent images allow us to provide graphics that look good whether the user has a low-resolution or high-resolution display and regardless of the size of the graphics they use. For icons and window borders, which the user may change the sizes of, this is a big win.
Additionally, SVG allows for animations and transformations that make it easy to provide very nice and organic effects in the UI. The implementation in Qt 4.1 is quite lightweight and very fast.
JL: Ah yes, there's been some comment about this. One Linux person I talked to made the point that if many things are ported to other OSes, why should they use Linux? I guess, where developers have chipped in on a Linux app or utility and it is then ported to, say, Windows, they might feel fairly grumpy. What do you make of this?
AJS: This is actually exactly the argument I made in my blog last year; that blog entry was then picked up by Slashdot, and from there it went all over. So I am in a nontrivial way responsible for that meme. And I still agree with it.
On the other hand, there can be real benefits to porting our libraries to other operating systems. The primary one is that we get more developers using our technology, which then translates into more applications for KDE. If people use our libraries to build their applications, then we can more effectively address the issue of desktop application availability on Linux and other open source operating systems. But many developers, particularly those producing closed source software, are not willing or able to write software that they can't deliver to people on Windows.
We will not be porting things like Plasma to Windows, however. In fact, Plasma (as an example) will not be portable because we will have hard requirements on X11 in it.
So, while there will be many KDE applications that won't be portable to Windows, our libraries will be available to Windows and Mac OS X developers to easily write powerful and portable applications. This will allow us to increase our developer pool and increase the number of applications available for KDE on Linux.
JL: Does KDE have a stated policy?
AJS: Not at this point, no.
JL: Is there any kind of target date for a beta?
AJS: We're aiming for a release in fall 2006. We don't have any firm timelines yet.
John Littler is chief gopher for Mstation.org.
Return to the Linux DevCenter.
Copyright © 2009 O'Reilly Media, Inc.