Robb Beal reveals this in his interview of John Siracusa [via daringfireball.net]:
Linux on the desktop needs a group of smart, dedicated people in positions of power who care about making a GUI that everyone can use and enjoy. Basically, they need the modern-day equivalent of the original Mac team. But that's not enough. On top of that, they also need an ego-maniacal dictator who will rule with an iron fist, crushing all opposition in the pursuit of The One True Linux GUI. This person has to have the power to change anything about Linux: the directory structure, the executable format, kernel features, the driver model, anything and everything. This team needs the total commitment and support of the entire Linux development community. Then, after a few years of initial work followed by several years of refinement, they might be in the ballpark of Mac OS X (assuming they don't make too many terrible mistakes).
Of course, none of the above is likely to happen in the world of Linux. But you just can't get to something like Mac OS X without a big effort by a lot of people pursing a single, coherent goal, and--most importantly--a willingness to change anything. No sacred cows.
Right now, the Linux community values "diversity" too highly to ever get a single, consistent GUI, let alone a good one. At the same time, it holds on doggedly to its (often ancient) Unix-rooted traditions and conventions. Finally, it's hard to get a really large group of Linux developers to do much of anything beyond a single "project." A GUI is not a "project." It's the whole OS from the user's perspective, and it must be from the creators' perspective too or it will fail.
Basically, I think Linux as an institution is allergic to a good, consistent GUI. Their priorities are reversed. They want to build a GUI on top of an OS. If they want to compete on the desktop, they should be building an OS to power their GUI. Of course, they don't have a GUI, which is problem number one.
They need to think of what they want the user experience to be, and then design a system that provides that experience. Period. This is so basic that even Apple forgets it from time to time. If you don't know where you want to go, you're never going to get there.
I think some semblance of "a Linux GUI" will arrive in bits and pieces over many years, but it will be almost impossible for it to compete with Windows, let alone Mac OS X, in terms of consistency, polish, and purity of purpose. Price, on the other hand... ;)
As for building the next Adobe or Macromedia, I don't think that's a very interesting goal. Both of those companies exist at the pleasure of Microsoft. If their businesses were broad enough for MS to want them, then MS would take them eventually. The only valid "grand plan" for dreamers just out of school in this day and age is to do something that makes MS irrelevant. That's pretty hard to do, but it's the only way to be anything other than a smart little rodent hiding among the rocks today's software market.
s response hits the nail on the head as far as I am concerned regarding one reason why Linux will never succeed as a desktop OS.
