.comment: Disco Night at the Old Folks' Home - page 2
Reeling in the Years
But the eeriest item in the magazine--I mean goosebumps spooky--was William F. Zachmann's column, "The OSF's CXI: A Thorn in IBM's Side."
What's more, H-P and Digital Equipment Corporation would be porting the thing, under the supervision of the OSF, to AT&T Unix and System V.
Zachmann noted that IBM wasn't especially happy dealing with other people's architectures, and had in fact been pretty much dragged kicking and screaming into Unix. "IBM began developing Unix products only when Unix became a requirement to win government contracts.... The result is that IBM supports standards no more than the minimum deemed necessary to respond (or at least to appear to respond) to market demands." The net effect of the new GUI would be that people could get the advantages of the IBM desktop without having to deal with IBM; soon, there could be Unix apps that might obviate the need for, oh, say, IBM mainframes. And they might even be easy to use!
(For an extra weird coincidence, I discovered that on the same page as Zachmann's column was an advertisement for a software company named Helix.)
Hmmm. IBM and H-P and the desktop. I could swear I recently heard something about this . . .
Actually, the intervening 11 years have made a difference in this case: IBM is no longer so much striving to keep what it has as it is trying to regain what it has lost. The National Weather Service is running Beowulf clusters to predict the weather. There is circulating in government a proposal that the feds give their official endorsement to Open Source. The market demands a minimum response, or at least the appearance thereof. IBM is touting Linux on its big iron. And of course IBM is joining with Sun, H-P, and others to develop a--what? why, yes--Common Unix Desktop. (They're not calling it that; the image of a cow placidly chewing, while in some ways appropriate, is probably not the kind of dynamism the companies wish to impart--besides, CUD might already be trademarked by Gateway.)
Speaking of Sun. Same magazine, in a three-page advertisement in which it implies that IBM and its OS/2 are everything but good, it sings the praises of Unix on its Sun386i systems. With such a system, the ad says, you can do currency trading figures under UNIX, update your Lotus 1-2-3 spreadsheet under DOS, make a Harvard Graphics slide show under DOS, paste the results of all of them into a Unix desktop publisher, and email it. "Far out, you say?" the ad queries.
Which is where the full length of the last 11 years hit me. The old money is starting to take an interest in Linux, and no matter what you think of it, it's a little pathetically comical, like disco night at the old folks' home. Or midlife crises, when men who ought to know better do things like buy motorcycles and marry beautiful, much younger women (though in at least one case I know of--me--this has worked out pretty well).
Yet it's probably more than that. Deep down inside, even the once ultra-straight IBM has undergone some fundamental changes. I remember a meeting one evening in 1994 that I arrived at directly from a sailboat I'd been racing for the last few days. I had two-years' growth of hair, two weeks' growth of beard, and was wearing ratty grey sailing shorts, a teeshirt commemorating New Zealand's Steinlager II sailing team, and boat shoes with no socks. The meeting was heavily populated by IBM employees, one of whom asked me during a break, without even a hint of sarcasm, "So. Are you an IBMer?" (I wasn't. But she told me that in some departments IBM had come to resemble Apple in the old days.)
The motives, I think, haven't changed much, and, really, there's nothing wrong with that. Companies exist to produce things from which they can generate revenue, hoping that some of it is profit, and from that they pay employees who then do things like eat and sleep indoors and buy computers.
But the methods have changed a lot. And that, I guess, is why I'm a little concerned about companies coming in to help Linux arrive at a standard desktop. The Open Source community can learn a lot that's good from its elders, but it can learn a lot that's bad, too. Organization of big projects and the establishment of useful standards can be a good thing; sticking it to the other guy isn't.
Of course, it's always possible that the big, established companies will learn some things from us, too: responding to user needs in a meaningful way comes to mind, as does producing hardware that doesn't dictate the choice of operating system.
That would be real progress. Not as much progress as putting the shop manual for old British Seagull outboard motors online, but progress nonetheless.