said it once before but it bears repeating

It’s hard to maintain a sense of what counts as novel. I run into this all the time when trying to come up with posts for TechDirt: the struggle to construct an interesting perspective rather than a bland restatement of what should be obvious to any casual observer. Yet, often as not when I fail and post something that seems like a vanilla rehashing of well-trodden ground, the comments fill up with people who are confused or shocked or in violent disagreement over what I thought were settled issues.

Basically, I need to get out more. On the internet, I mean.

Or I at least need to recognize that there might be some merit to repeating myself. With that in mind, let’s talk about software patents — partly because of this newfound resolution, and partly because I just can’t help myself.

So: Matt put up a post pointing to Tim’s review of Math You Can’t Use, a book which explains why software patents are a bad idea. Yglesias’s commenters are being surprisingly dense about this. Let’s try this one more time:

When ever anyone owns property (real, personal or intellectual) there are always those complain the whole system of ownership is a drag on them. They are right. They are also blind to the bigger picture. Failure to protect legitimate property interests has bigger problems, long term, in that there is no incentive of the property owner to invest and develop the property if he cannot obtain a return on his investment. It is no surprise that there is a strong relationship between protection of property rights and long term economic growth.


Last I looked, I didnt think that those countries who had nationalized property for the “national interest” were doing so well, economically.

But of course physical property transfers are zero sum. That’s not true for intellectual property transfers — copies of ideas are free. Screaming “commie!” at a copyright reformer is incoherent and always has been.

[W]hat I see is that A gets a patent, B finds it inconvenient, and argues that the whole system is broken. The main argument presented by B (and friends) is that it makes life and business difficult for B.

Well… yeah. But this is a good argument. Again, the glaringly obvious: the enforcement of intellectual property laws (and thereby the concept of intellectual property itself) is about costs and benefits. What cost shall we impose on society by restricting the value-generating copying of ideas in order that new ideas will remain plentiful and of high quality? We stop people from copying songs so that singers will keep singing. We stop people from copying drugs so that chemists will keep formulating. We stop people from selling counterfeit Mickey Mouse paraphenalia so that crazed fascist sympathizers will continue to dream up beloved children’s characters and, one day, be able to afford to cryogenically preserve their bodies.

Clearly, there is an equilibrium to be sought. We need new drugs, but we do not want aspirin to cost $10 per pill. The thing is, the last few years have demonstrated that for many areas of creative endeavor we have massively overestimated the compensation that must be provided in order for society to continue enjoying plentiful ideas. The success of the open source movement and the continued fecundity of the music industry, for example, clearly show that we, the greedy masses, can help ourselves to considerably more free intellectual output than we’d thought without endangering the whole system. We’ll all enjoy the benefits — as well as the savings to be had by not spending as many of our resources on enforcing IP laws.

Of course, how far this can be pushed is up for debate. Other fields — publishing, filmmaking, drug discovery — may not be as adaptable. In some cases I’m optimistic that the associated capital costs have or will fall to the point where creator compensation can be safely diminished in order to benefit us all. But now we’re getting into the weeds. Clearly there ought to be a negotiation around each IP-producing field to determine whether the benefits provided by restricting the use of its output outweigh the value lost by imposing that restriction.

And for software patents, the situation is extraordinarily clear: it’s very hard to find a software author who thinks that patents have contributed anything useful to the process of innovation. In fact, most developers think that patents have exerted a significant retarding effect. If that weren’t the case — if the ideas produced and protected under the software patent system were the sorts of novel and valuable contributions that the system is designed to encourage — one might expect to have heard about them.

Well, which ones have we heard about? The three most famous, I’d say, are the Amazon 1-Click patent, which concerns users buying items with a single mouse click; the LZW patent, which involves the compression algorithm used in the GIF file format; and the Eolas patent, which is about a web browser’s ability to embed other programs, like Flash movies, Java applets or ActiveX controls, into rendered webpages.

The first thing to realize about all of these is that they’re pretty stupid. The Amazon patent is famous because it’s such an obvious idea that its legal protection inspires disbelief. Perhaps more compelling is the Eolas patent. But again, the notion of having an application deputize another program to render content is pretty obvious; in fact, there was prior art that many thought should have prevented the patent from being granted. It was not an idea that needed protection in order for it to spring to life.

But the LZW patent is probably the most emblematic. LZW is a compression algorithm that was granted patent protection. Compuserve baked it into the GIF file format without realizing it was protected, and GIF attained widespread use. Then the owners of LZW came forward and started asking for payments. Whoops! Almost immediately the PNG format sprang into being — it turns out that LZW wasn’t all that important; there were plenty of other ways to solve the problem, many of them free. But by that point GIF was baked into enough applications that for many there wasn’t much choice but to knuckle under to the demand for royalties.

This kind of so-called “submarine” patent is the real threat. It’s a tremendous boon to our society that a creative individual can start writing valuable software with a few hundred dollars’ worth of hardware and an internet connection. Software patents add “… and an attorney on retainer” to that list. It’s simply not possible to know if you’re in violation of a patent without conducting an expensive search. And even that’s no guarantee.

Imposing significant costs in order to subsidize things that don’t need subsidizing is wasteful and dumb — particularly when the relative costs and benefits are this far out of whack.

3 Responses to “said it once before but it bears repeating”

  1. James Hare

    YES YES YES!!
    This post should be required reading for every patent examiner, every member of congress, the President and many other people I can’t think to name.
    Software patents are teh sux0rz. And if Microsoft really has patents on Linux they oughta put up or shut up.

  2. Phil

    Look, I’m extremely sympathetic to the idea that perhaps software patents are a bad idea, but I’m not sure I really the buy the second to last paragraph.
    In a nutshell, it seems like you’re saying that it’s extremely cheap to develop very valuable software, no use mucking about making it more expensive. That can’t be right. Pharmaceutical development is extremely expensive, yet it’s ok to add even more costs by creating legal protections that have to be designed around? Isn’t the real argument here is that “valuable: software developments are hard to monetize, hard to protect, and perhaps don’t return enough to make a retainer worth the trouble?

  3. Ken Kennedy

    Well-spoken, dude. These arguments just keep rising from the grave…we gotta just keep cutting them down. Damn, it’s tedious at times.

Leave a Reply