The Desktop Linux Question, Re-Re(-Re)visited

My colleague Jonathan Salem Baskin clearly isn't one to shy away from a good controversy. How, he asks, can Linux realize its full potential without a robust desktop edition as part of its diversity?

My original answer to this was cast as an answer for another, more basic question: Does Linux need to succeed on the desktop? I voted no, simply because I felt Linux was already a success, and using the desktop as a yardstick for further success on top of that was misleading.

Part of the reason I took this POV was because at the time the prevailing notion of "success on the desktop" went something like this: either knock Windows out of the top slot, or make something that will be at least as solid a choice for the desktop as Windows or the Mac. I felt that was a distraction from what was Linux's best strength: its flexibility and mutability. Sure, someone could sit down and engineer a Linux-based desktop environment that would be appealing, but they shouldn't try to do so with the explicit aim of making Windows, etc. irrelevant.

Since then, I've tried to stay mindful of how this situation might change. And I've found that what matters most, I think, is not that Linux succeed on the desktop per se but that they learn many of the same lessons as those who have come before them on the desktop. The art and science of the UI, mainly -- something both Microsoft and Apple have invested countless hours and dollars in, and applied back to their products from the top down.

Linux hasn't done anything like that -- or, rather, the various Linux distributions / desktop environment creators have taken a "let's try a little of everything and see what sticks" approach when it comes to UI innovation. The end result has been a well-intentioned hodgepodge, with everything from the maybe-it's-a-Mac flavor of gOS to the KDE 4 disaster. It's not that there's no consistency between the approaches -- that's a red herring -- but that each approach seems bound to make mistakes already made and transcended elsewhere. KDE 4 and the Compiz desktop-effects bundle are two of the worst offenders in this regard: they're nice to look at, but they not only add nothing to productivity but may in fact actively detract from it.

I've said in the past that Linux UIs don't need to follow the Windows model slavishly, that they should feel free enough to innovate. I'm now of the feeling that if they don't at least understand why Windows and the Mac made the choices they did, no amount of innovating will make up for that if all they do is embrace things that were discarded for good reasons.

It's not success on the desktop that Linux needs, but a sense of what success on the desktop would bring them. That would consist of feedback from users -- non-technical users -- about how their UIs behave, and what they expect to happen vs. what their users expect. It's easier to get that sort of thing if you already have a huge percentage of the desktop market, so rather than let this devolve into a chicken-vs.-egg argument, I'll put it this way: The best substitute for success on the desktop is some sense that they need not reinvent the wheel, and badly to boot.

I'll close with an analogy -- maybe not a perfect one, but I'll share it anyway. I once knew an aspiring fiction writer -- call him "Alan" -- with a staunch resistance to being mentored. He didn't want a teacher; he wanted to be his own teacher, to make his own mistakes and make the discoveries he needed to make on his own, because that way they would truly be his.

I pointed out that this seemed to stem more from his image of himself as a self-made man than anything else. People who shirk the help of others for the sake of "making their own mistakes", I noted, wind up having so much of their time eaten by repeating other people's mistakes that by the time they get around to learning something, they're exhausted and dispirited, and the lesson is lost on them because they have no idea how to pick up on it.

Social philosopher Paul Goodman put it this way: " ... starting from scratch, without literary tradition altogether, writing and reading are imbecile and trivial; ancient errors are tiresomely repeated; platitudes are taken for ideas; hard won distinctions are lost; useful genres have to be reinvented, like reinventing shoes or learning to boil water." (Emphasis mine.)

He, too, was talking about writing, but if you ask me, those lines about shoes and water are pretty universal.

"Alan", by the way, has to my knowledge never finished anything.

InformationWeek Analytics has published an independent analysis of the challenges around setting business priorities for next-gen Web applications. Download the report here (registration required).

Follow me and the rest of InformationWeek on Twitter.