I use a Linux desktop every day. Specifically, I run SuSE Linux Professional 9.0 with KDE 3.1.4. On top of that, I use Mozilla 1.5 for my Web browsing, OpenOffice 1.1 for my word processing and spreadsheets, and Gaim 0.74 for my instant messaging needs. For those times I need a Windows program, usually PhotoShop and Quicken, I use NeTraverse's Win4Lin to run them on a Windows 98SE virtual machine.
For me, it works great. But I know Linux like the back of my hand. So do most people who use Linux desktops today. For most corporate knowledge workers, though, the Linux desktop still isn't good enough.
It's not that most company employees couldn't pick up Linux. They could. This isn't the 1980s when the big *nix desktop choice was between the Bourne shell and the C shell. Today, KDE and GNOME are both fine graphical user interface choices.
The real Linux desktop adoption issue for businesses is whether the Linux desktop is enough of a win over legacy desktops to cost-justify the move. Today, for most companies, it's not.
That's because corporate IT departments still need to put the pieces of the desktop together for their users. Do you use KDE or GNOME? OpenOffice, StarOffice or just go straight to the whole Java Desktop System? Mozilla or Konqueror?
Linux fans love to debate these issues. CIOs, however, don't want a debate. They just want to make a decision that will work for all their users, and that comes with strong external support.
Many Linux people—who love to argue about the technical virtues of their favorite GUI—still don't get this. Trust me, guys, businesses don't even care.
A prime example of this kind of conflict between programmers who want to be right and business users' needs is the fuss between Bruce Perens and KDE supporters over Perens' UserLinux. Perens gets that businesses want a best-of-breed desktop Linux that doesn't require much installation or support effort. The KDE fans don't get this point at all. They insist that UserLinux must include KDE as well as GNOME.
Wrong. Most businesses don't want to maintain and support two entirely different GUIs. For them, supporting both KDE and GNOME is a complete waste of time and money.
If the KDE fans really want to do their cause credit, they should stop trying to shoehorn KDE into Perens' GNOME-based Linux and build their own Debian-based Linux with a KDE interface. After all, this is Linux. Anyone can build his own distribution. Then it would be the business customers, not programmers, who'd make the final choice between KDE and GNOME-based desktop Linuxes.
Next page: Linux desktop edges closer.
Despite such fusses, the Linux desktop is getting closer—thanks to two developments.
The first is that Microsoft is making upgrading Windows and Office an ever more expensive proposition. It was one thing when you could sit pat as a business user when, flaws and all, Windows 98 was what you already had and Microsoft still supported you. Those days are almost gone.
It's very simple. You can upgrade a machine that runs Windows 98 to run Linux cheaply. But to run XP Pro or W2K Pro, you must not only upgrade your software, you must buy a new PC to run it on. Throw in the costs of switching to Licensing 6, and even the most ardent of Microsoft shops now have good reason to at least consider a Linux alternative.
This isn't just me mouthing off. On Dec. 30, Israel stopped buying Microsoft office programs in favor of open-source alternatives. The Israeli government did this because Microsoft's programs were too expensive. I expect you're going to see a lot more stories like this one in the coming months.
The other reason is that Linux desktops are maturing. Sun's Java Desktop System is a good complete operating system/office suite package. I think Perens is on the right trail with UserLinux. I have no doubt that Novell/SuSE/Ximian will have something tasty cooking up for corporate users by 2004's third quarter. And I'm happy to see Linux desktop operations like Xandros not only building a good desktop, but including administrator tools so that technicians can remotely set up new Xandros desktops.
I've said it before, and I still believe it: 2004 won't be the year of the Linux desktop. But I do think the foundations are being dug for 2005 to be that year.
eWEEK.com Linux & Open Source Center Editor Steven J. Vaughan-Nichols has been using and writing about operating systems since the late '80s and thinks he may just have learned something about them along the way.
Commentary 
