Speaking in London last week, Richard Stallman, founder of GNU, argued passionately against the legalisation of what he calls "software idea patents". The core of Stallman's argument is that if companies are allowed to patent software ideas, big business will ride roughshod over the smaller players, and the free software movement will be effectively strangled.
This, Stallman warns, is the fate awaiting Europe if the current draft of the directive on computer implemented inventions is passed unchanged. He argues that different fields need different approaches. Patenting might be appropriate for the pharmaceutical sector, but he believes fervently that it is not appropriate for software.
This won't be the end of free software, he says. But it will kill off much of the community. "Sure there'll still be some of us left," Stallman noted. "But that's like saying its OK if millions of Africans dies of AIDS because there are loads of Africans left."
In the US, where software patents are granted routinely, the independent developer faces a huge challenge if he wants avoid infringing a patent, Stallman says. Consider the huge volume of patents filed, the difficulty in understanding what exactly has been patented in any one application, the fact that patents pending are not always in the public domain, and so on.
When the developer infringes a patent, he has three choices: avoid the idea, try to buy a license, or overturn the patent. No-one is obliged to grant licenses, and they can name their terms and it wouldn't take too many deals requiring a slice of gross sales to sink a product.
Big companies can ride out these kinds of difficulties by brandishing their own portfolio of patents, and signing cross-licensing deals. They can also afford to pursue their patents in the courts.
The draft European legislation dealing with computer-implemented inventions was proposed to provide some clarity across the Union as to what should and should not be patentable. However, it is wide open to interpretation. At one extreme, it can be read as only allowing patents for software that supports a physical process, but at the other it could be applied much more strongly. Some believe it could even be argued in court that it does allow software business methods to be patented.
Against this political backdrop, Stallman's message is an important one, so it is a real shame that it gets clouded by his choice of analogy. There is little doubt that allowing patents on software will have a devastating impact on the free software community, and good reason to believe, based on the current situation in the US, that it will hurt smaller companies working in the field.
Likening this impending doom to the AIDS crisis in Africa is counterproductive, and merely allows pro-patent groups to label Stallman, and by association the anti-patenting movement, as a crackpot. It also makes it more difficult for politicians to stand up and be counted as being against software patents.
Fortunately, this was not the only analogy he offered. A better example is that of music: the component ideas in software are very similar to the building blocks of music. In the same way that a scale here and a chord there do not a symphony make, there is more to a word processor than a spell checker and a couple of font definitions.
"Yes, patenting rewards ideas, but it also bolloxes the system that produces them", Stallman said. "If in the 1700s the government decided to allow musical idea patents, by the time Beethoven was composing, he would have had a harder time writing something that didn't infringe a patent than he would writing his symphonies."
Stallman argues that like music, progress in software is dependent on the ideas that have gone before. The creative and innovative part is not the ideas themselves, necessarily, but the particular combinations that make up the whole. Patent the component ideas, and no one can write anything new. ®