The Death of Many (If Not Most) Software Patents?
patentlyo.com
patentlyo.com
I can't think of any company or engineer I know of that would stop writing software if they couldn't patent it, or of anyone that's learned anything useful from reading a patent filing.
The most important parts of software are already protected by copyright and trade secret laws, and patents have absolutely no place there.
finally a good change of direction for the USPTO
The client (#12 in the suit if you're curious) called me in a panic telling me to pull down the site, so all its users are screwed. I've also run a free wishlist site at GreedyMe.com for several years which wasn't named in the suit, but if it had been, I couldn't financially justify trying to fight it.
PageRank was pretty revolutionary. I'm on the fence as to whether that should be patented. Amazon's 1-Click and Channel Intelligence's "list storage" are obvious methods, and I can't fathom how any lawyer or businessman can submit a patent for such nonsense with a straight face.
So you just move your data repository to one machine and your data processing to the other, and you have your patents back.
I really just don't understand how something like this happens. I guess "The Law" is a massively backwards-compatible system that makes computer architecture anachronisms look mild.
However, the inventions which have been patented recently are often obvious and trivial. Amazon's one-click patent is not some amazing secret that we have been let in on by the patent system.
As compared to the situation today, where Google's patents allow them to be completely open about how their systems work, and we all get to read the source code to Microsoft's software because it's protected by patent?
Companies already keep their nontrivial inventions as secret as they can, because they know that the patent system is useless for its (supposed) intended purpose. [1] The system moves at a glacial pace relative to the industry, it costs too much to use, its results are unpredictable to the point of randomness, it's filled with reams and reams of invalid chaff, and frankly software itself is too abstract, mutable, and easily copied to be reliably protected by patent. There's a good reason why the inventors of patents didn't let people patent written works. Those folks weren't dummies.
No, software patents are a tool for trolls, nothing more. Every invalidated software patent is a victory for mankind.
[1] If, OTOH, you believe that the intended purpose of software patents was to allow deep-pocketed corporations to patent-troll their smaller competitors to death... the system is working precisely as designed.
Of course, to a large corporation such mosquitoes don't matter much. Most patent-troll companies can be paid to go away -- that is, in fact, their entire business model. And giant corporations get a relatively good deal in those negotiations, because they've got the money for a defense and their own giant collection of dubious patents to deploy in countersuits.
It's a cost of doing business. Microsoft has to pay their legal department anyway, because even without software patents I'm sure that the company is sued dozens of times a day.
It's true that, once in a blue moon, a software-patent holder will have a good enough case, and a stubborn enough attitude, to at least appear to seriously threaten the interests of a large company. But I think it's pretty rare. And appearances can be deceiving when parties are arguing over money -- bluff and bluster are important legal tactics.
This is much too strong, I think. I'm a co-founder of a small biotech startup with a single patent to its name. We had trouble enough raising money even with the patent. Do you think that I, or anyone else, would have put up with all the pain and the long hours in the lab if we new that Agilent or Biorad could swoop in at any moment and take our markets away from us using their marketing clout?
>> If, OTOH, you believe that the intended purpose of software patents was to allow deep-pocketed corporations to patent-troll their smaller competitors to death... the system is working precisely as designed.
Sometimes, no doubt. But Borland and that small storage company both cleaned Microsoft out for over $100,000,000 apiece, and MSFT's market cap is a bit larger.
A patent is a limited term legal monopoly which effectively creates a legal partnership between a government and, say, a three person startup on the one hand and, oh, Google on the other. Without the patent it's you and your buddies against Larry and Sergey and you're just bringing a knife to a gunfight. Since this site is largely devoted to startups, this would seem to be an important point.
I think the real solution is to shorten the terms of software patents to, say, 7 years and tighten up the meaning of "obvious to one schooled in the art". That would immediately reduce the patent infringement search space while still giving a small company time to build a first-mover market advantage.
Do you think that I, or anyone else, would have put up with all the pain and the long hours in the lab if we new that Agilent or Biorad could swoop in at any moment and take our markets away from us using their marketing clout?
I'm sure that folks reading this site -- the ones whose products could be cloned by Google or Yahoo at any moment; the ones whose earlier companies were crushed by Microsoft -- have an entire orchestra of tiny violins to play for you.
I used to build multiphoton microscopes for a living. The story of the multiphoton microscope is amusing. Once upon a time, Bio-Rad held the exclusive rights to the original Cornell patent. Their competitor, Zeiss, read the patent and decided to argue that it covered only the use of sub-picosecond laser pulses for microscopy. So Zeiss came out with a competing system intended for use with 1 picosecond pulses. Of course, if you found yourself in the lab late at night and you happened to turn the knob on your laser that shortened the pulses to 80 femtoseconds, your Zeiss microscope presumably continued to work pretty well, but with a big improvement in image brightness and depth. I assume that Zeiss refused to be held responsible for that improvement -- after all, the scope wasn't intentionally designed for short pulses, because that would be Legally Dubious.
Naturally, the whole thing went to court for years. The issue was settled when Zeiss bought out Bio-Rad. See here:
http://listserv.buffalo.edu/cgi-bin/wa?A2=ind0401&L=conf...
I can't tell you who won or who lost, there, but here's my conclusion: If a company the size of Agilent or Zeiss or Biorad wants to violate your patent, they'll just do it. All you were doing, in those nights in the lab, was trying to improve your negotiating position in the resulting court case.
(That, BTW, was a traditional hardware patent, not the software kind. If yours is a software-only patent... good luck with that.)
See my poll: http://news.ycombinator.com/item?id=253318
In the world without software patents, there's no way of knowing how it works, neither exploiting it. Although of course there's the advantage of using independently made similar inventions without fear.
The latest style is to use a mix of patents and trade secrets anyway, meaning that even if you wasted your time reading patents, you still wouldn't have the full picture.
The good part is now everyone has to compete for users with what users wants. So a feature on product A can be quickly copied to product B by its competent competitors. At the same time, product A's creators has to keep on moving on or face replacement. While workers in company for product A may have tougher time in life.
Software is rather unique in that it has been granted all of patent, copyright, and trade secret status.
I firmly believe that software patents are counterproductive to advancing the industry and should not be held as valid. The legal developments detailed in the article look like great progress in encouraging innovation in the software industry.