.comment: On Writing About Linux - page 2
The Paradox of News
At the site of that long-ago Indiana plane crash, I watched a reporter for a Louisville station do a perfect, straight-faced "standup" (when the reporter stands in front of the camera, the scene in the background, and gives his account) and then, when the 16mm camera his cameraman was using was no longer running, burst into tears. (Today, the reporter would burst into tears on-camera and probably win a Peabody for it, but that's a subject for another article, in another publication.) I myself have covered events ranging from the new pilot who flew three of his high-school friends into a thunderstorm, crashing minutes later, killing all of them; I was the second person on the scene. I have watched a man who swerved to avoid the little car that pulled out in front of him and who therefore hit a bridge abutment, trapping him in his truck, burn to death, unable to be rescued by the firemen who had been working on it until the gas tank took fire. I have written about those things. I was not, in writing about them, endorsing them, nor was I trying to sell newspapers. I did hope that in writing about them, I might cause people to think a little about those events: about the precarious nature of life, about how the overwhelming confidence of youth, aimed improperly, can be a fatal affliction, about how when you're not careful driving you can get killed or cause the death of the fellow who swerves so you won't get killed.
It is a fact that no report is utterly devoid of influence drawn from the reporter's opinions. When quotations are taken, they're never all that was said. They are the reporter's estimation of the portion of what was said that is most important, that most crystallizes the speaker's point, that is most illustrative. A different reporter might well make different choices. A careful reporter, knowing he favors one side, might well overly compensate and write a piece that accidentally favors the other. And today (more grist for that other article sometime, someplace else) opinion is a greater part of reporting than ever before.
My article in LinuxToday was labeled "News and Analysis." This means that the news was stated along with the reporter's -- my -- estimation of what it might mean.
This column, though, is an opinion piece, and all of the above is to describe some of what goes into reporting, and reporting on Linux -- subjects on which it seems that all people are experts until they actually try to do it, at which point the winnowing out is quick and merciless.
So here is what I think about the story I covered yesterday.
I'm not even slightly conversant with the GNOME development community, though I am very conversant with people who are informed observers of that scene. It is indisputable that something unhappy took place there, and the prevailing view is that someone of value has departed that community in anger. Was it a big deal? There's no firm rule for measuring such things, but it was a big enough deal that Alan Cox, who no one accuses of being less than a serious fellow, felt the need to weigh in strongly. I may have erred in noting, almost as a footnote, that there was a problem with a GNOME maintainer gaining CVS access; if I did, I apologize.
I know more about the KDE situation, having followed its development, developers, and developers mailing list for years. In that time I've had communications with both Mosfet and Waldo Bastian. If vague memory serves, I think I've been on the opposite side in flame wars involving both of them. I've also had very friendly exchanges with both of them. They are both tremendously talented people for whom I have enormous respect.
As Waldo has pointed out, he did not personally return any of Mosfet's code to the CVS tree, nor did he personally yank Mosfet's CVS access. Still, he is responsible for the KDE-2.2 release, he agrees with the decisions to do those things, and it was within his power to overturn those decisions. George Bush did not personally attack Iraq, but when it's said he did everyone understands what's meant. I think Waldo understands that digressing into the details would be to miss the point, though I'm glad he later provided those details.
And I understand that Waldo has a difficult task; on paper it's the kind of thing that looks impossibly difficult: Hundreds of developers contributing big things and small ones to a huge project that somehow needs to be tied together, tested on a multitude of disparate distributions and platforms under a wide variety of compilers and ancillary libraries, with the gamut of computing equipment, by a specified time, without benefit of the kind of employee hierarchy found in commercial software efforts. (March that development model over to IBM and see how far it gets you!) Yet it works, and usually without too much fuss beyond a little slippage of the release date.
On the other hand, my sympathies lie with Mosfet. He has cut his own path, followed his own muse, thought out his own work beginning to end, and coded it. He has advanced the technical palette available in KDE. He does his best work alone. His best work is very good indeed. He has given a great deal to KDE. He has the quirks that attend creative endeavor -- the literature is rife with accounts of brilliance accompanied by attitudes ranging from odd to truly weird, but in all cases it involves nonnegotiable belief in what one is doing.
It requires a sense of independence. Like most reporters, I have insisted that my name be removed from stories that editors have altered into oblivion. (The Directors Guild actually has a formal process -- if a motion picture is deemed by the director to have been changed such that it is no longer his, the director credit in the film goes to "Alan Smithee.") Like many, I've resigned in anger from paying jobs. So I cannot help but feel for the guy, and hope that he has not been fastidious in his bridge burning.
I also see demonstrated what I've long thought would come to be a real concern: Mosfet bemoans his lack of control over his own code, the fruits of his own labor, his own vision. In his commentary that appeared briefly on his web page, he talked about his work being forked but not renamed, thereby causing confusion in that he intended to maintain separate and maintained versions, available from him. That is no doubt part of it. That cannot be all of it.
Most creative people take pride in their work. They are troubled by the thought of it falling out of their control, of being modified thoughtlessly, of being taken in directions other than the one dictated by the original vision. No architect has succeeded in figuring graffiti into his design; the Mona Lisa would not be improved by a moustache.
Linus is still in charge of the kernel. It could be forked, of course, but that's a practical impossibility.
There are lots of big projects that are community efforts; many are very good, and many are destined forever to suck. Likewise single-developer projects.
The question becomes what happens when those of a single-developer personality participate in community projects. I think that most everybody has a level where personal pride in one's work becomes the governing factor; I believe that this figured in the GNOME dispute and I know it did in the KDE one.