The X Desktop (Editorial)

Here is something to think about. A friendly war is being waged over what your X desktop will look like in the years to one. Hit the link below to read about the two major contenters, KDE and Gnome, what I think about them, and why they need to work together.

The Battle for the Unix Dekstop is on. It's pretty much accepted that nobody cares about CDE anymore, and the future lies along another path. But what path is that? Two have emerged now. The major contendors for our future are Gnome and KDE.

There are many similiarities between the two standards. Both feature good looking, extensible widget sets. Both provide a unified look and feel accrossed applications. And both have an army of passionate supporters.

But the differences are where the problems lie. KDE has been criticized for it's use of Qt, Troll's popular widget set used in numerous applications. The problem is that Qt isn't "Free" in the RMS definition of the word. People beat on K for this. Realisticly, Qt is 'Free Enough' for all but the most hard core user, but many people are opposed to the tool kits we rely on being anything less than Free. I don't think that this alone is reason to shelf KDE, but it looks like public opinion is headed that way already.

The second major opposition to KDE is the similiarities that it has with Windows. Hard core unix users often use unix to get away from the interface that KDE has tried so hard to emulate. The similiar widgets, and the start-button like cascading menu however are a good start for new users seeking handholding as they adapt to a new OS. I think it's important that the veteran Unix users realize that, and just remember that this interface is for the newbies.

Gnome doesn't have license issues- the G in Gnome stands for GNU, and the whole package is under the GPL. Gnome is the younger of the desktop options, and it shows. While K has a full Window Manager and numerous applications showing off slick integration, Gnome is all rough edges. But these rough edges are being smoothed at an amazing rate as GTK gets closer and closer to one point oh. And don't forget the premiere Gnome Application, The Gimp. A new desktop won't drive installation for many systems, but the Gimp will. And since you already need GTK, why not throw the rest of Gnome on too? This reasoning will drive Gnome forward.

Recently RedHat opened up their Advanced Development Labs and announced that one of their major projects was advancing the X desktop. To that end, they snagged Raster, one of the talented programmers largely responsible for The Enlightenment Window Manager. Soon after, Raster ported E's image library (imlib) to work with Gnome's GDK.

But now KDE has released beta 3, and while the stability is a bit lacking, and there are holes you could drive a family sized sedan through, it is a cohesive product that is getting very close to it's goal of producing a standard X based desktop that looks good, is easy for a new user to adapt too, and is reasonably customizable and extendable. Some distributions have decided that K is already strong enough to ship.

But what now? I wish I had a simple answer here, but I don't. I see 2 groups of people working on amazing software. But I don't want this whole effort to result in the need for every modern unix system to install 2 more sets of libraries to run a few more applications. And I don't want to see programming efforts duplicated by multiple people around the world.

I don't mind diversity, it really is what makes Unix the cool thing it is today, but this time we need to work together. Both KDE and Gnome have a web browser. This can't be efficient! Both will have status indicators, irc clients, panel applications, task viewers and text editors. And each of these programs will have a couple of talented coders hacking away on them.

So I don't know what the answer is. Perhaps a set of C++ wrapper classes that allow KDE applications to compile under GTK? There must be away to bring together these 2 wonderful products in a way that allows both of them to accomplish their goal of a better UI for X in less time.

What do you guys think?