KDE is a mature, functional GUI for Linux and Unix that’s popular with seasoned users and newbies. However, KDE doesn’t get much love from enterprise Linux vendors, who seem to be gravitating toward the rival GNOME environment.
The move away from KDE and toward GNOME is not a positive development. Both environments have strengths, and were better served by healthy competition between them.
The latest KDE snub came last month, when open-source mover Bruce Perens announced that he’d chosen to exclude KDE in favor of GNOME from the forthcoming enterprise-aimed, community-led UserLinux distribution.
GNOME is the default desktop for Red Hat’s high-profile Linux products, as well as the environment that Sun has selected for Solaris and for its Java Desktop System. SuSE Linux has been KDE’s most prominent supporter in the enterprise arena, but SuSE’s recent acquisition by Novell has left the door open for a potential switch. Novell also purchased the desktop Linux company Ximian, which is a leader in the GNOME development community.
It’s puzzling why KDE has had trouble securing a strong foothold in the enterprise. Besides boasting an active developer community and as large a user base as GNOME’s, KDE has corporate-friendly features that GNOME does not, such as its Kiosk lockdown mode and support for remote desktop sharing.
In addition, many users find the KDE interface more similar to Windows than that of GNOME. That similarity can make migrations from Windows easier and is one reason why Linux distributions tailored for new users, such as Lindows and Xandros, tend to use KDE by default.
The most commonly cited problem with KDE is the way the framework on which it’s built, Qt, is licensed. Qt is a product of Trolltech, which distributes the framework under a dual-license scheme. The X11 version of Qt is free for open-source projects but requires a per-developer license when used to create proprietary software. The community is split on the importance of this issue, but this is the reason Perens cited in explaining his choice of GNOME, which is built on a framework, Gtk+, that may be used to create free and proprietary software without royalties.
But Qt is also one of KDE’s biggest strengths. Qt provides good development tools, and it works on more platforms than Gtk. It supports Windows, Linux/ Unix, Mac OS X and embedded Linux.
Next page: Divergent design philosophies
Divergent design philosophies
There’s also the matter of divergent design philosophies. KDE tends to err on the side of extra functionality and configurability while GNOME takes a “less is more” approach. For instance, KDE’s file manager, Konqueror, does a lot more than GNOME’s Nautilus, but Konqueror’s many buttons and preference screens can seem bewildering. Konqueror also serves as KDEs default Web browser, and the environment ships with its own productivity suite, KOffice. This makes for better integration among KDE applications. Depending on your point of view, KDE is either monolithic and bloated or integrated and complete.
After writing in a column about my preference for KDE, I spent the next several months running GNOME as my desktop environment. While I missed certain KDE functions and found myself spending more time at the command line, I found GNOME to be more attractive-looking than KDE. Also, I have found that Konqueror as a browser and KOffice arent quite on par with Mozilla and OpenOffice.org. Since I prefer the Gtk-based Gaim and Evolution for instant messaging and e-mail, respectively, I don’t take advantage of the integration benefits of using K applications across the board, anyway.
It may be that enterprise Linux distributors find GNOME’s simpler design a better base from which to build and differentiate their offerings. But its important to remember that while GNOME has racked up some recent enterprise vendor wins, both groups are moving forward quickly, and its too early to write KDE off.
Although the UserLinux distribution will use GNOME, there’s a project already under way to push KDE as an interface alternative.
So, bottom line, let’s not limit our choices by deciding either/or. For now, both/and is a better answer.
Senior Analyst Jason Brooks can be reached at [email protected].