More on software patents

Software patents are still in the news. The European patent directive continues to lumber forward, despite overwhelming opposition from groups on all corners -- excepting the big software companies that will win big if sofware patents become law in Europe as well as the U.S.

Not surprisingly, Richard Stallman has no kind words for software patents. In response to Stallman, fellow ZDNet blogger and Microsoft employee John Carroll, however, does have a few things to say about the benefits of software patents.

He hits on three alleged benefits for software patents, which I'd like to touch on here. The first is boost to small business. Carroll says that software patents can be "the bedrock upon which small companies are founded" and then says "that matters, at least if you think corporations matter."

Thus far, I've seen no small businesses propelled forward thanks to software patents -- unless you count "businesses" that specialize in patent lawsuits. I'll return to the discussion about whether corporations matter in a bit.

Carroll also argues that patents are necessary because of the costs of research and development, and compares the software industry to the medical drug industry. There are a couple of reasons I don't buy into this argument. The costs of software development, in general, pale in comparison to the costs of drug research. The equipment necessary, clinical trials and years of development for drugs are far more expensive than developing your typical application. Sure, Microsoft may be approaching that kind of R&D funding for Longhorn as a project -- but Microsoft wouldn't patent Longhorn -- they patent specific pieces of the project, which don't come close to adding up to one significant medical invention.

As our patent system works now, a single programmer can come up with an idea and implement it in a very short time and apply for a patent -- assuming they have the funds to work their way through the patent process, of course. In fact, the cost of getting a patent may be higher than the cost of developing the "invention" in the first place.

Finally, Carroll argues that software patents are a benefit because they encourage dissemination of knowledge. That's certainly a worthy goal, but ultimately worth very little in the software patent debate. The reason I say that is because the knowledge provided in a typical software patent does little to forward the science of software development.

Most of the time, it's quite possible to implement something that's been patented without any of the information published in a patent. Consider the patent granted to Microsoft for "Time based hardware button for application launch" granted last April -- a method for clicking buttons to start applications. We needed documentation from Microsoft to move the field of computing forward in this area? I think not. What patents are very effective at, in the software industry, is blocking innovation.

As has been discussed many times before, fighting software patents costs a lot of money -- something that Microsoft and other large corporations have in abundance, and small startups are lacking. Even if they are sure to win, many companies can't afford a protracted legal battle over software patents when they're trying to establish themselves in the market. Is Microsoft's button-clicking patent more likely to A) disseminate knowledge, or B) be used against a competitor as part of Microsoft's massive patent arsenal? (Actually, it's probably C) collect dust and serve as a reminder of silly patents.)

Let's return to the issue of whether corporations matter. Carroll posits that "Stallman is no fan of corporations," and suggests that Stallman has "shaped this debate in religious terms and turned it into a battle of absolutes."

Unfortunately, this is a battle of absolutes. Remember, the issue that prompted Stallman's piece is the European patent directive, which threatens to introduce U.S. style software patents in Europe. That is, in fact, a fairly black or white proposition. If one supports software patents, then it's a big plus to see that Europe accepts the same kind of software patents that are already prevalent in the U.S. For those of us who do not support software patents, it's a big negative.

Whether Stallman is in favor of corporations or not, I have no idea. However, I see software patents tilting the playing field too far in favor of a small subset of corporations, and ignoring the interests of small companies and FOSS developers and users -- which is basically the rest of the population, given the ever-widening use of open source and free software. It's not whether or not corporations matter, it's whether their interests should be paramount and whether benefits to corporations are benefits to society in general. In this case, I see the "benefits" of patents being limited to those corporations, and not something that benefits the population at large.