As any regular reader of this blog will know, I'm an every day KDE user. After a history of using SuSE and Gentoo, my Linux distribution of choice for the past year or so has been Kubuntu, the KDE flavour of Ubuntu. Well, after a few discussions at this month's Birmingham Linux User Group meeting, which coincided with the new release of the Ubuntu family, the Gutsy Gibbon, and having spent quite a bit of the evening strongly advocating KDE, I've now decided to give GNOME a try.

For those not up on the particulars of modern Linux-based desktops, GNOME and KDE represent two quite different approaches to a fully fledged desktop system. The UNIX world has always been modular, and though the graphical windowing system itself is provided by X Windows, much of the look and feel is provided by the particular window manager or desktop environment you choose. The two most fully featured and mature desktop environments are KDE and GNOME.

A bit of history is probably in order here. Many people moving over to GNU/Linux based systems spend a while wondering why there are several different options, and why the community doesn't just pool its effort into one really good system. There are many reasons why that doesn't always happen in open source development, but the KDE/GNOME case is particularly interesting.

KDE was begun by Matthias Ettrich in 1996, in an attempt to create a consistent and easy to use desktop environment for UNIX systems. It was immediately a very popular project and gained developers fast, the software progressing quickly. However, many in the free software world (including myself) had concerns about KDE, since it was based on the proprietary QT widget libraries, owned and developed by the Norwegian company Trolltech. Of course, the GPL does not allow free software to be linked with proprietary software, and hence the GNU project could not support KDE. Debian subsequently dropped the system from its distribution.

A number of responses came about from this move. Listed in no particular order, the first was an attempt to create an alternative, free version of QT, called Harmony. Secondly, many keen KDE people created the KDE Free QT movement, an attempt to persuade Trolltech to relicense QT under a free licence, and subsequently to guarantee its availability should Trolltech disappear or abandon QT. Thirdly, the GNOME project was founded; an attempt to create an alternative desktop environment from scratch, free from any involvement with proprietary code or reliance on commercial developers.

Those who know where I stand on free software would expect that I would have been a supporter of GNOME from the start, and back then they'd have been right. I was doing my undergraduate degree at the time, and the university computers included both GNOME and KDE. So, I used this early version of GNOME for a while on both them any my own PC.

GNOME has always felt different to use than KDE; a bit like the difference between AmigaOS and Windows 3.0, both of which I'd been used to switching between previously, so after a year or so of using KDE, GNOME took a bit of getting used to. My initial reactions were quite positive; it looked good and was quite fast, though much of its functionality appeared to exist in framework documents rather than the software I could see on my screen. Still, it allowed me to achieve a consistent look to the desktop (something previously lacking on Linux really) and the look was highly configurable. It looked better than KDE.

However, over the next couple of years, to me, GNOME stagnated. KDE, on the other hand, came on leaps and bounds. KDE's file manager (kfm back then) and subsequently its lightweight web browser (konqueror) appeared technically superior to Netscape Navigator, which the GNOME systems hadn't been able to supplant. But by far the best feature of KDE back then was the interoperability of the applications. Items could be dragged and dropped between virtually any application, and they would know what to do with them. Even the early implementations of what is now known as 'kparts' technology allowed programs to be run inside others when Microsoft Office was still struggling to achieve this properly.

Then in 2000, the freedom issue seemed to disappear. Trolltech released QT under the GPL, resolving the issues about distributing modified code, which their own QPL created. By this point, an agreement was also firmly in place with KDE to ensure that if Trolltech folded or abandoned QT, then the code would be released under a BSD style licence.

By this time, both GNOME and KDE were progressing very quickly, though KDE remained consistently a year or two ahead of GNOME in terms of features and stability. Since the new millennium, both systems have been very usable, though each retaining a distinct look and feel. At this point I was already very comfortable with KDE and its ever expanding suite of programs. I found GNOME a bit like KDE before I'd upgraded it. Added to that, GNOME lacked (perhaps still does) a uniform way of doing various tasks (such as right clicking on a menu bar to configure it). Its file dialogs were chunky, and it just didn't have as many features as KDE. I stuck with KDE.

The choice of desktop system by the main Linux distributions at this time was fairly evenly split. Many had no obvious bias, whilst Redhat used GNOME, SuSE favoured KDE, and both contributed heavily towards the development of their respective desktops. Debian now included KDE again, and a host of other systems lay on each side.

Then, in 2003 Novell bought SuSE, and subsequently also brought in GNOME founder Miguel de Icaza and his company Ximian, who were responsible for a large chunk of GNOME development. It was no great surprise, therefore, when Novell's SUSE (note the large 'U') announced that GNOME would be the preferred desktop for their Linux distributions, with which they hoped to compete with Redhat in the enterprise market. Luckily, SUSE still do contribute towards KDE, though the announcement did send ripples across the KDE world. Controversially, Miguel has now led the charge on Microsoft compatibility through the Mono project, and despite the guy having done great work on pushing open file formats in recent years, I've heard frequent criticisms of his style of "starting" projects, only to leave others to pick up the pieces and make the thing work.

Arguably the most popular desktop Linux-based system today is Ubuntu, itself a remodelling of that stalwart of the Linux world, Debian (rather excellently named after its founders, Deb and Ian). Ubuntu however, was also firmly behind GNOME, and despite the fact that KDE and XFCE based flavours of the distribution were later released, the lion's share of development and use remains focussed on GNOME. With all this extra focus, surely GNOME has now caught up.

Given the number of distributions now preferring GNOME, its no surprise that its user share has also increased. A hunch tells me that more people at my Linux User Group use GNOME than KDE, and this month we were treated to a demo of GNOME-based Ubuntu's latest features. It looked good. Ubuntu, being primarily developed for GNOME, tends to be slightly quicker off the mark with new functionality for the GNOME flavour than the KDE one, but how much? So, as someone who has now used KDE since version 1.0, and only dabbled with GNOME occasionally, perhaps it is time for me to take a closer look.

So, my "project", if you will, is to spend a month using GNOME rather than KDE on my home PC. I've installed Ubuntu Gutsy Gibbon from a fresh install, and there isn't a KDE application in sight. I'm going to attempt to get to know the GNOME way of doing things. I aim to compare GNOME and its applications' approach with the good KDE functionality I know I like. I also hope to discover some new stuff too.

In a month's time, I will write about my experiences, and whether it's tempted me to make a more permanent switch.

After all, KDE and GNOME (contrary to many mailing list debates) are not enemies, just rival siblings, willing each other to do well. And it's good to get to know your friends better, right?