We interrupt this broadcast to repeat the message that Software patents are evil.
While I am not an expert on the matter I hold the following basic facts on patents to be true:
Proponents of patents would have you think that patents are a natural extension of property law, and by analogy a natural extension of common rights we have under traditional Western ideas of law, but that is not the case. Quite the opposite is true.
The history of Western power is the history of the free exchange of ideas
Patents run counter to that idea
Patents introduce artificial scarcity. Artificial scarcity is good for the few who can exploit it to their advantage, but bad for society. Unfortunately, there's a lot of money in running a scarcity racket, so there's a lot of pressure to introduce more schemes.
There are legitimate uses for patents, but software isn't one. Patents were never meant to introduce the notion of property, but to protect innovators from free riders. This makes sense in industries that have a skewed distribution of cost, so that almost all the costs in innovation are accrued before any profits are made.
The opposite is true for software
In general, the cost of innovation is low compared to the cost of exploiting the innovation commercially. If this wasn't true you simply would not have many upstarts. New pharmaceutical companies don't ship product while having less than 10 employees. New software companies do that all the time.
What allows free riders in innovation to operate is a situation where the cost of 'second invention' is marginal compared to the cost of 'first invention'. This is true for the pharmaceutical business, where most of the cost accrues in proving that the innovation works at all, and only the first innovator has to do that.
Software re-innovation is expensive
This is so because of the observation above, that the cost lies in the commercial exploitation, not producing the actual patentable innovation. A reinnovator still has the exploitation cost.
Copying software innovations is simply not something that happens a lot. Instead the common scenario is the one described in the landmark book The Innovator's Dilemma. New, small, agile companies introduce new products often significantly simpler than the dominant product in a category, but cheaper - catering to new market segments that the dominant player is incapable of feeding. Only gradually do these new companies evolve to the same quality as that of the market leading offering.
Patents kill this dynamic where technology, through reinnovation by new players, finds new applications in new markets. This dynamic is not a problem, it is a good thing. Without it, we wouldn't have PC's running Windows but would still have an 80% IBM market share and would still be working on mainframes.
It takes one look at the market for operating systems to see the chilling effects the speed of the legal system could have in stifling the natural speed of the software evolution process (and the SCO case is not even about patents).
There is another skewed economy at work in patents: The cost of applying for a patent and the chance of getting it, even if the claim is later found to be frivolous, is significantly lower than the cost of challenging a patent. That introduces an incentive to mass-patent broad ideas. As more and more people do so, the patent process becomes increasingly poor, incapable of handling the increased accumulated complexity of the claims made.
Patents are slow, software is fast.
If you slow the software process by running it through the legal process, you're not only introducing a capital cost manageable for the few, you're also dropping the software innovation cycle from it's current speed to something several times slower than that.
Is I say, I'm not a legal expert in any way, but I think the evidence for these basic claims are visible in plain sight all over the place. I plan to expand on them, and document them more tightly in the near future.
Look here for ways to say no.Posted by Claus at October 21, 2004 11:07 AM | TrackBack (0)