"If customers come in with a PC disk, tell them we use Macs. If they have a Mac disk, tell them we use PCs. If they have both, tell them we use Linux." - "PopCopy" training video, from Chappelle's Show DVD, Season I.
In the three years since this skit first aired, Dave Chappelle achieved superstardom, made millions on DVD sales, freaked out and ran to away to South Africa, and then came back to tell the tale. But Linux on the desktop is still pretty much in the same state of obscurity it was in 2003.
Linux desktop market share still hovers at around 5% to 7% of all PC operating systems. Part of the problem with making desktop Linux as serious a technology as it is on the server is fragmentation. Linux distributions are still as all over the place as spilled ball bearings. And not only do software makers have this to think about when writing code, but many distributions themselves offer multiple desktop environments to confuse matters further - the old KDE vs. Gnome debate.
Efforts to solidify Linux desktop technology are underway, but even this is becoming fragmented. Earlier this month, code was released from The Portland Project, an effort backed by the Open Source Development Labs to create common interfaces and tools that allow KDE and Gnome desktops to use the same software. Meanwhile, the Free Standards Group is set to unveil its Linux Standards Base 3.1 this week at the Linux Desktop Summit. The new LSB will also outline how KDE and Gnome will work together.
"There are a lot of people working on the [Linux] desktop stuff. I think it's a worthy project," says Bob Gatewood, CTO of AthenaHealth, a Watertown, Mass., healthcare applications hosting firm that relies heavily on Linux and open source in its data centers - but nowhere on its corporate desktops. "I just worry that because there's no central planning, that talent and resources sometimes go to the wrong place."
Gatewood cites the myriad desktop software projects, kernel tweaks, obscure hardware driver ports and other projects going on all over the open source community as parts of what might be holding desktop Linux back. He also cites what is arguably the most successful open source project as an example that desktop Linux could follow.
"One thing that's real interesting to me is the way the Apache foundation organizes itself," Gatewood says. "Someone should write a Ph.D. thesis on that model - a total meritocracy. Having some governance and method of control on making decisions, and directing resources to the highest-value developers is important."