A spectre is haunting the software industry-or at least the part of it controlled by Bill Gates and his fellow moguls. It's called the Open Source movement; its basic proposition is that proprietary software (the kind created and sold by companies like Microsoft) is a flawed idea. For some Open Source activists this is about technology; they are convinced that proprietary programmes are less reliable, less stable and more bug-ridden than programmes which are communally owned and created through the cooperative efforts of hundreds, even thousands, of dedicated hackers. Other Open Source adherents believe that their software is not only technically better than anything produced by Gates & Co., but also that it is ethically superior. For them, the notion that programmes should be "owned" by individuals or corporations is as odious as the idea that humans could be owned and traded as slaves.
The term "Open Source" is a recent euphemism, coined by the pragmatic wing of the movement to ease acceptance of their ideas by big business. They felt that "free software" sounded too frightening. In fact the movement dates back to the early 1980s when Richard Stallman, an MIT researcher, developed the idea of free software. "Think of free speech, not free beer," he says. Users ought to be free to modify programmes to meet their own needs. The only way to make that possible is to distribute programmes in their original code, rather than the binary form in which proprietary software comes. This is why my Microsoft Internet Explorer is not free in Stallman's sense: although Microsoft gave it to me gratis, I am not free to alter it, because I only have the binary code. The source code is locked in Bill Gates's safe.
Stallman's other contribution was to invent a new type of licence which gives one the right to alter a programme and pass (or even sell) it on to others-provided one also passes on the same right to them. He calls it "copyleft," to distinguish it from standard methods of protecting intellectual property.
Until recently, most people-even in the industry-knew little about the "free software" movement. Its adherents were regarded as pony-tailed relics of the pre-Microsoft era. The intellectual property rights embodied in proprietary programmes had created global corporations like Microsoft and Sun Microsystems, and made huge personal fortunes. Free software belonged with caftans and the Grateful Dead on the scrapheap of history.
And now? Suddenly the industry is scrambling to board the free software bandwagon. Scarcely a day passes without a big computer supplier-Dell, Compaq, IBM even-announcing that it proposes to install free software on its products. So, what's changed?
Answer: Linux, a free programme which is a clone of Unix, the system originally developed by AT&T and now owned by SCO, a Californian software company. Linux was created in 1990 by a Finnish student named Linus Torvalds, and immediately posted on the internet. Over the next few years, an enthusiastic community of hackers developed Torvalds's original little system into one of the most stable and reliable pieces of software ever created. And they did all this without any financial reward, for Linux was-and remains-a free product. It can be downloaded from the internet, purchased on a CD for the cost of the disk or bought, bundled with installers and other utility software in special "distributions" sold by companies such as Red Hat for less than ?50.
Linux was an open secret in the academic research community for several years before the business and media worlds discovered it in 1997. What brought it to wider attention was the leak (on the internet, naturally) of an internal Microsoft discussion paper revealing that the company viewed Linux as a threat to its key strategic product-the network version of the Windows operating system (now called Windows 2000, and still behind schedule). The paper also revealed that Microsoft viewed the methodology and resources of the Linux community as in some respects superior to its own. On the grounds that if Bill Gates took Linux seriously there must be something in it, the world woke up to the disturbing thought that the correlation between price and quality might no longer hold-at least in software.
This was followed by the realisation that all the software which runs the internet-the most resilient and fastest-growing communications system ever constructed-is free. It was developed in the same collaborative style as Linux, is entirely in the public domain, and is reliable. The Apache web server programme-which runs half the world's websites-is free. The 1,000 or so servers which power Yahoo!-the most popular internet search engine-all run on free software. The list goes on.
Open Source software is so good because it is debugged in the crucible of intensive peer review. When the source code is available, people can poke around in the programme to see how it works, and fix it when it breaks. Operating systems like Windows 2000 are highly complex, with billions of possible behaviour modes. Even Microsoft staggers under the load of debugging them. (And it shows-as any heavy user of a Windows PC knows to his or her cost.)
The Open Source community has thousands of people all over the world who spend much of their time testing codes to destruction and fixing bugs when they find them. The community can throw far more expertise at the task than any company could muster. As Eric Raymond, an Open Source evangelist, once said: "Given enough eyeballs, all bugs are shallow." And Netscape's recent decision to release the source code of its browser has generated much discussion about whether other companies should harness the collective IQ of the internet to debug the code-monsters they have spawned.
Many of the free software movement's most prominent members are fierce individualists, reluctant to suffer gladly anyone with a sub-200 IQ. Getting such people to cooperate is normally like trying to herd cats. Yet in the movement they work productively together. Raymond has described the community as a "high-tech gift culture" in which status is determined not by what you own but by what you give away. What decides a hacker's place in the programming pecking order is the quality of the code he donates to the collective project.
The free software movement raises some big questions. Does it provide a new method of producing high-quality software, or does it apply to only some programme categories? What does it mean for the economics of cyberspace? How can free software coexist with hard-nosed profit-maximising companies? And will the movement, riven by tensions between Open Source pragmatists and free software fundamentalists, fall apart under the pressures of success? Open Source works in practice; will it work in theory?
