ZDNet's Dana Blankenhorn has an interesting post suggesting that uniformity in the Linux desktop market -- i.e., fewer distributions -- may be the key to finally seizing marketshare from Windows.
Blankenhorn suggests that perhaps Ubuntu, with re-selling endorsements from Dell and Acer, may be taking the lead as THE Linux desktop distro, but ends with an open-ended lament that desktop Linux still seems fragmented, with Red Hat having no serious skin in the game.
I'd suggest that the real tipping point for desktop Linux will come when the applications that users are accustomed to are available for desktop Linux. This is true for most consumer tech users; it's 10 times more sure in the business world, where any change -- no matter how rudimentary -- means expense.
I've spent the last few weeks playing around with Ubuntu, and found it to be pretty useful and friendly. I'm no support tech, so I can't speak to the learning curve for support staff in dealing with Linux desktops, particularly in a heterogeneous environment, where the first wave of Linux client adoption will need to gain a beachhead -- businesses simply aren't going to pitch Windows unilaterally. But I can say that most users should be able to come to terms with the trashcan being anchored in what Windows users would call the task bar.
However, a more daunting curve exists for desktop applications, which are far more complicated than the OS GUI, at least from a business user's point of view. However intuitive I might find Ubuntu's interface, there is no way I'm going to undertake getting a handle on the GIMP image editor -- it took me long enough to figure out my low-end Photoshop drills, thanks.
I have, however, written a handful of posts in WordPress on our Ubuntu test system. Obviously, as a Web-based application, every WordPress session runs identically regardless of what OS a host system happens to be running. I have nothing new to learn.
So, my guess is that desktop Linux will begin to gain traction when all applications are forced to run in a browser or, more likely, in OS-independent runtimes that support local data access and offline functionality. Adobe's Apollo quickly fell out of the headlines, but I still think it's the sign of things to come.
Within the next decade or so, businesses will expect to license software through the SaaS model, and as Linux and Mac nibble at the Microsoft OS stranglehold (and nibbling is all you should expect, at least in the U.S.), there will be no financial incentive for companies to tie their software to a given OS's code base. When line-of-business managers no longer sense a cost with a move to desktop Linux, IT will be at liberty to calculate its own ROI on maintaining client machines.
In other words, desktop Linux will begin to take marketshare from Microsoft when the desktop OS becomes an absolute commodity, like Web browsers are now.