Last week, Bob Metcalfe predicted that Linux would fizzle against Windows. His reasoning? Open-source ideology is balderdash and Linux is 30-year-old technology.
As editorial director of LinuxWorld, it's not surprising that I disagree with Bob's conclusion. And to follow a precedent set by Bob, I'm prepared to eat my column if it turns out I'm wrong. If Bob hasn't changed his mind by January 1, 2001, that's what I'll do.
I'm confident Bob will realise he has fallen into a classic Microsoft trap. When Microsoft knows it is losing a battle, it attempts to refocus the argument to something it can win. Consequently, Microsoft isn't predicting it will win based on stability, price, support, or even applications. It wants people to believe that the fate of Linux is bound to the universal adoption of open-source ideology.
Unfortunately for Linux, Microsoft has a lot of unwitting allies in the open-source movement. These allies too have decided to make Linux a battlefield in a war between capitalism and free software. According to Microsoft, once the profit motive kicks in, Linux is dead. If anything, profit is why Linux can't lose.
Linux will continue to erode Windows momentum for the same reason that Windows stole the limelight from Unix. And this is the same reason Unix stole momentum from minicomputers and mainframes: cold, hard cash.
Unix was created between 1969 and 1970 for a couple of popular and inexpensive minicomputers of the time, the Digital PDP-7 and PDP-11. Technically, Unix was an open-source operating system during much of its early development because AT&T distributed the source code freely to universities. In 1978, AT&T began to charge licence fees for the source code. AT&T decided to take Unix commercial in 1979.
That's when Unix began to fragment. But contrary to revisionist history, the fragmentation of Unix didn't slow its momentum. After all, the key Unix vendors such as Sun Microsystems and Digital Equipment Corporation weren't really selling Unix. They were selling hardware. To them it was cheaper to port Unix to their hardware than it would have been to create an operating system from scratch.
The demand for these machines persisted because they were the least expensive in their class. It was cost-effective to use and support these machines because the universities were churning out people with Unix expertise.
As a result, Unix did so well that by 1987, Hewlett-Packard and IBM had to port Unix to their platforms simply to remain competitive in the advanced workstation market.
Unix didn't begin to hit the skids until IBM PC clones started shoving advanced workstations off the desktop. DOS rode in on the wave of the PC's success, followed by Windows. The decision to bundle DOS and Windows vs anything else was a no-brainer. The choice boiled down to roughly $10 plus support costs vs porting and supporting Unix or some other OS.
The equation didn't change enough to make a difference when Microsoft jacked up the price with Windows 95. And compared to the alternatives, Microsoft could easily sweep the server market even with a higher per-copy price for Windows NT.
But then Linux entered the picture. And it did so just when Microsoft had to convince the US Department of Justice that it has competition. That allowed Linux to enjoy an initial surge of momentum. That's all it needed.
To vendors, the battle is now between Windows 2000 and Linux, and it comes down to hundreds of dollars per PC vs nothing. Zip. Nada. Linux also represents a fighting chance to software vendors who have been or may be crushed by Microsoft. This, folks, is a formula for success if ever there was one.
As for your other objection, Bob, no doubt the reason you fail to see 30 years as a period of refinement is that you are using Windows NT as your yardstick. Trust me. In the rest of the world, technology actually improves over time.
Speaking of which, I'm about to use a nicely refined 26-year-old technology to file this column. You may have heard of it. It's called Ethernet.
Nicholas Petreley is editorial director of LinuxWorld (www.linuxworld.com). Reach him at firstname.lastname@example.org, and visit his forum at www.infoworld.com.