Back to article
.comment: If Not Now, When?
May 23, 2001
The Linux community is forever blathering on about "free as in speech," which in practical terms means that people who don't own it ought to have rights to other people's code.
Time, then, for an object lesson in real free speech. Two of them, actually. Last week in this space I wrote of my belief that standardization, not of the content of Linux distributions but of where those distributions put things, would be good for the widespread adoption of Linux, especially on the desktop, and would make life far easier for application developers. My good friend and onetime technical editor (Practical KDE) Bob Bernstein did not agree. Boy, oh, boy, he did not agree! He committed his disagreement to text, and that text appeared the following day on Linux Today. Bob and I have some pretty fundamental disagreements about the direction Linux ought to head, and when he finally unloaded he went as far as to allude to Nazism in characterizing my proposal. He will disagree with me this week, too. And he is still my friend. That is how free speech works.
In the intervening week, my point has moved closer to being moot. Mandrake, it appears, has not made a lot of money selling desktop Linux. If reports are accurate, things are pretty bleak at Mandrake. This leaves, by my count, three distributions with a chance of surviving: Red Hat (unless its money all goes to defending investor lawsuits, which is a sad possibility), SuSE (with the field thinning rapidly, SuSE has the most to gain from a mass migration of mere users to Linux), and of course Debian (and its close commercial relative, Progeny). We all expected some shaking out and some specialization. Storm Linux is gone, and Caldera has decided there's no money in chasing the desktop user. Seldom does a week go by that the rumors of TurboLinux's passing don't resurface, and from that we can at least draw inference that its fiscal health is not robust.
Indeed, with reports of Mandrake's situation coming out on Monday, my editor and friend Kevin Reichard offered the view that Linux is dead as a desktop operating system. And now it is my turn to disagree, at least somewhat. That, friends, is free speech.
I disagree with Kevin not in his view that things look bad for Linux as a desktop operating system but in his view that the cause is lost. I don't think it need be. Indeed, the potential for Linux to become a major player on the desktop has never been better. Whether that potential will become reality remains to be seen. Let me explain the factors that I believe represent the brightest opportunity that Linux has ever had.What Would It Take?
History is full of evidence that the most technically profound way of doing things doesn't always win through sheer technical profundity. I don't think that even the engineers at Japan Victor Corporation, its inventor, would seriously argue that VHS is superior to Sony's Beta format for home videotaping. Indeed, Sony owned the market for a few years a couple of decades ago. But Sony used its position to keep prices high, and its licensing fees were such that other manufacturers found JVC's way of doing things more fiscally attractive.
Likewise, after putting out the first PCs with open architecture that could be copied by anyone, IBM developed its far superior Micro Channel Architecture. This it did not make open, choosing instead to license it to other manufacturers. Other manufacturers decided that EISA, VLB, and, later, PCI were cheaper and almost as good. There are few MCA machines around today -- I know Alan Cox has one, but then again I still have a Betamax around here someplace.
An outfit in Berkeley cranked out a nifty little GUI and productivity suite a decade ago called GeoWorks, and from a technical standpoint it knocked the socks off of the just-released Windows 3.0. But the company was slow in putting together a software development kit, so it never gained the momentum it deserved. IBM -- in a litany of things done wrong, IBM always gets at least a couple of mentions -- produced OS/2, which had features that no one else, nine years later, has equaled. It also ran 16-bit Windows applications in addition to its own native, multithreaded, really cool apps. Microsoft was at the time being extremely kind to application developers, while IBM, startled by the fact that actual users were migrating to OS/2 (there was actually a period in late 1994 when OS/2 Warp3 was the top-selling software in the U.S.), proved an unsteady partner at best. So developers decided that by writing for Windows they could sell to users of both programs -- Windows was at that time just a DOS GUI, not an operating system -- and there ended up being relatively few OS/2 apps.
The common thread here is that through mishandling, superior technology never enjoyed the success that one would figure it would achieve, looking at the specs .
It can't be argued that Linux isn't technically superior to anything coming out of Redmond. If technical achievement were the determining factor, it would already be the dominant operating system. It isn't, on both counts.
Now. There are those who argue against the widespread adoption of Linux, and still others who would argue against it if it would mean the slightest bow toward that goal -- arriving at a common file hierarchy standard, for instance. If you are among those, go read something else or spike a tree or something -- I have no time for you now. But if you think that there is reason to hope for widespread adoption of Linux on the desktop, read on.
We've been asking the wrong question.
The majority of computer users got their machines in a box, all put together, with speakers, and a CD reader, a big hard drive, and Windows. (And a 3.5-inch floppy drive. Does anyone else still keep a 5.25-inch drive in the production machine? Have you used it anytime this century? Me neither.) The machine came with various little applications, and maybe some big ones, and they mostly work. Fact is, for the home or small office user who doesn't try anything very exotic, Windows and its applications aren't all that unreliable. These people can go to the store and find shelf upon shelf of software and hardware that is designed for and will work just fine with Windows.
What would it take to get these users to migrate to Linux? That's the question we ought to be asking. (And no, the answer is not to make Linux just like Windows, which is what the angry mob always falls back on when, after about 15 seconds, their anger entirely replaces their already limited powers of reasoning.)
To get a user to abandon something that he or she is using, one needs either to capitalize on the user's dissatisfaction with the present product or offer something so vastly superior that it's worth the trouble.
Windows users seem largely satisfied. That may be about to change, and in that change is the first step toward Linux on the desktop. Microsoft's new licensing policies are going to anger a lot of people and inconvenience damn near everybody. They're designed to prevent counterfeiting their software, which is the only way it has achieved dominance in much of the world. In the far East, one buys Windows, any version, from a pirate for $5 per CD. In the Mideast, one doesn't usually even have to pay the five bucks -- not long ago I reformatted a hard drive, looking over its contents first to make sure I didn't have an ohnosecond, remembering something there I wanted to save, and I found an Arabic version of Word 2.0, put on the machine by someone when I was in Kuwait in 1993. But if having the CD with Windows on it does you no good, a substantial part of the world is going to be looking for alternatives.
As, truth be known, will a lot of Western businesses. Many small businesses, with, say, a five-node network, will find ponying up the bread for five licenses of everything a sudden and troubling expense. Ah, you say, so they'll just stick with what they have. Well, maybe. But if they have to exchange documents with customers, they will have to upgrade, because Microsoft has mastered the art of selling upgrades: They change the file formats. Do you really think anybody would have bought Winword in the last five years but for the need to read the new file format?
So the opportunity for Linux to make inroads on the desktop has never been better. That does not assure success, though. In some countries it could just as easily result in a wider divide between the haves and the have nots, as those who can afford it pay their tribute to Microsoft, while those who can't keep using what they have or simply decide against investment in a computer. And we mustn't overlook the likelihood that Microsoft's little system will get cracked; everything else they have has been.
But the potential for Linux to make its move is there.