Are the changes in the Longhorn road map really the "best chance ever" for desktop Linux, as some have suggested? Only if the difference in going from "teensy weensy" to merely "teensy" can be considered a big improvement. To the rest of us, either size is still just a speck.
Using this sort of measurement, one might even say Longhorns trouble improves desktop Linuxs chances by 100, 1,000, or even 1,000,000 times. Its just that youre starting with such a tiny chance that even a big multiplier doesnt make it much larger in practical terms.
And why are Linux and Longhorn supposed to be linked, anyway? I dont think people who are looking at desktop Linux have been sitting around befuddled by the choice. Linux vs. XP is on peoples planning radar, but to be thinking about Longhorn vs. Linux today gives you a better early-warning radar than the North American Aerospace Defense Command uses at Cheyenne Mountain.
Even if Longhorn ships in mid-2006, it will be mid-2007 before we see it in real quantity, late-2008 before it has a significant installed base, and perhaps 2011 before XP finally fades away. Thats if Microsoft hits its marks from here on and customers actually want what Longhorn manages to deliver. Now factor in time for WinFS and the confusion when XP starts getting some Longhorn features, and it looks like the fun just wont be stopping.
There are doubtless some places where desktop Linux makes sense. I think of these as places where people arent so much using a personal computer as they are a terminal or workstation. These are places like call centers, factory floors, and other situations where a thin client or terminal could work as well, and probably better, than something with a lot of local storage.
Linux is often an excellent choice for these applications and youd be wise to consider it. It is also, just to save the angry letters, a good choice for Web severs and database applications.
But Linux is not a good choice for user desktops. It is possible for a motivated user—whether by anti-Microsoft sentiment, a light wallet, or the desire to be a technological Flying Wallenda—to build a useful Linux desktop. Ive even built one myself.
Not that I enjoyed using it. The best Linux desktops are still beyond the skills of average mortals and enjoy very limited application support. How such systems could benefit from a Longhorn delay escapes me. Linux will remain the desktop of a specialized audience regardless of anything Longhorn does or doesnt accomplish. Longhorn is intended for the masses; well see if it delivers.
Now, its not that I have anything against non-Microsoft operating systems that causes me to poo-poo desktop Linux. What ought to happen is desktop Linux proponents should wake up and switch to Mac OS instead.
People who use Mac OS X and think about such things consider it to be "Linux done right." Of course, its Unix, not Linux, but thats a minor point. As Linux fragments, as it must, the distinction will become even more academic that it is already.
Mac OS X has a better user interface than Windows, better security and is more stable. Its all the operating system most users need. It doesnt have the application support Windows enjoys, but there are many more excellent commercial OS X apps than there are Linux apps.
If the desktop Linux people really cared about using a great operating system, they would stop trying to reinvent the wheel and rally around Mac OS X. It would be great if the contest were between two worthy operating systems, like Mac and XP/Longhorn. But Linux vs. Longhorn? Youve got to be kidding.