This week Google was granted a patent for "Social computing personas for protecting identity in online social interactions"; in other words, "a pseudonym could be presented as someone's name based upon their choices of who would see that name or their 'real' name." Sound familiar? It does to me. Two months before the patent was filed, I wrote here "Suppose Google let Plus users define which of their Circles see their real names, while others can only see their nicknames?"
I know, I know. You don't patent ideas, you patent implementations. I was interested enough to fight my way through the legal wording and try to work out what they're actually talking about. (Which was not easy, incidentally, even though I'm an experienced software engineer; patentese is incredibly hard to read.) There are a few nifty ideas in there: "one-time anonymous personae" and "a single persona with multiple users." Cool. But there's really no breathtaking insight or brilliant solution. The whole thing just seems like a very competent solution to the problem of online pseudonyms, one much like that which any other small group of good programmers would have designed.
I don't mean to come down on Google. Their patent policies are much better than those of most tech companies. But I bet most Google engineers along with the overwhelming majority of software engineers in general would agree with the following statement in a heartbeat: most software patents are total bullshit.
The other day I was working on a pet coding project and came up with a clever little hack that solved a nagging problem. And then I thought: I wonder if that's patentable. The terrifying answer is maybe. If no one else had documented, or (more likely) bothered to document that particular solution to that particular problem, I might be able to establish a personal monopoly on its use for the next 20 years, more or less. How crazy is that? I could even come up with a fancy-sounding name: "System and method for partitioning a single index into multiple independent search spaces." But it's something that any other good engineer would have worked out in fairly short order.
Mark Burnett recently discovered that "every imaginable aspect of password selection, authentication, storage, and recovery seems to be covered by one or more patents." Simply "checking for common or weak passwords" is covered by no less than seven different patents, almost all of which look like pure anyone-could-and-would-have-thought-of-this bullshit (though there are a couple that look more interesting again, not all software patents are nonsense).
Meanwhile, a company named PersonalWeb has filed suit against GitHub and Rackspace with yet another trove of bullshit patents. Ironically, these 'inventors' are so technically brilliant that they don't seem to quite understand the complex concept of "cloud hosting," as they're suing Rackspace only because GitHub is hosted on their servers. You have to laugh. Because otherwise you'd weep. (Remember when RIM had to pay out $600 million because of five invalid patents? Good times.)
PersonalWeb is, of course, incorporated in East Texas, because the courts there are little more than a speedy rubber stamp for even the most egregious of patent trolls. It's both infuriating and depressing. Great swathes of the American patent system have become little more than a pathetic and tawdry playground for rent seekers. And before anyone starts shouting that I want everyone to be free to steal inventions, here's everything you need to know about today's patent system on the back of a napkin. I give you the Tabarrok curve:
The saddest thing about all this is that (almost) everyone in the tech industry already knows it. Writing about how patents are bullshit, again, actually feels lazy. We're practically overflowing with solutions to the problem. But a sizable number of enormously powerful interests are heavily incentivized to keep the system in its current spectacularly broken state. I mean, unless you count full employment for patent lawyers and patent examiners (who have become the cryptic and draconian high priests of the tech world, à la the Wizard of Oz) as the highest and most noble goal of our species. Alas, nobody has a sufficiently powerful countervailing incentive to fix it, unless you count the incentive of, you know, doing the right thing but when was the last time that was a genuine factor in any decision made by Congress?
There are signs of light on the horizon. StackExchange's AskPatents.com. The recently launched Internet Association. But given the strength of the entrenched patent interests, I suspect the only way for the tech industry to change the patent system is to put together a massive campaign-contribution fund and outright buy some reform. The horrifying extent to which the American political system can be skewed by cold hard cash is hardly a smaller problem than the broken patent system, but maybe for once that sow's ear can become a silk purse. Otherwise, I fear we're just doomed to an endless Sisyphean treadmill of industry-wide patent wars and gridlock.
So I can't blame Google. Again, they're better than most, only engaging with the patent monster to the extent that such engagement has been forced on them. But next time, guys, just throw a cite my way, OK? I mean, what the hell we all know it's bullshit anyways.
Image credit: the Tabarrok curve, Alex Tabarrok, Marginal Revolution.
No hay comentarios:
Publicar un comentario