4 ms·

This presents a problem for legitimate patents - researchers and others who develop technology would only be able to profit from it by bringing it to production

This presents a problem for legitimate patents - researchers and others who develop technology would only be able to profit from it by bringing it to production themselves, rather than selling or licensing the technology to others and continuing to do research.

JoshTriplett 14y ago
Can you write any description that reliably distinguishes between a patent troll and anyone else who does not use the patent themselves in production (commonly called a "non-practicing entity" (NPE))? If not, that should tell you something. (Not that practicing entities can't resemble patent trolls too, as evidenced by the numerous ongoing smartphone lawsuits.)
- karamazov 14y ago
  The issue with patent trolls isn't how they use patents - it's the patents that they use. There's nothing wrong with a researcher suing a factory over a patent the researcher holds that the factory is using without license. There's also nothing wrong with the researcher selling the patent to a lawyer, so as to avoid dealing with the legal system, and that lawyer then suing the factory. This works if the patent was issued for a meaningful advance. In fact, the researcher may not have been interested in doing the research (or at least in revealing his methodology) without patent protection. If the methodology were not revealed, other people would need to spend the time to develop their own process, when it would be easier for them to pay the researcher and then use his work. The issue with software patents is that the methodologies covered are trivial, and the time needed to develop your own method is trivial as well. Because the patents are granted for obvious solutions, many people end up violating the patent without ever having seen it; this is why patents are not supposed to be granted for obvious developments.
  - AlisdairO 14y ago
    To roughly recall a comment someone made to me on HN a while back, if it's profitable to think of ideas and do nothing with them until you can sue someone else independently coming up with the same idea, there's a problem with the patent system. The whole point of patents is that they're for ideas that it would be unlikely that someone else would come up with on their own.
  - Zak 14y ago
    The problem I see with software patents is that they're usually attempts to patent ideas, but ideas aren't patentable. The implementations of most of the ideas in question are obvious and therefore not patentable. I don't think that we, as a society should decide to allow ideas to be owned, so I do think we should crack down on trivial patents.
    - runako 14y ago
      >> we should crack down on trivial patents. I think everyone is in agreement on this, but the definition of "trivial" is the sticking point.
- Avenger42 14y ago
  What about a distinction between people who license a patent to only one company (e.g., a university researcher creates a fast-and-accurate HIV test, the university licenses it to a company to produce & market it) versus companies who hold a patent and try to collect licensing fees from every company under the sun, most of whom have no connection to another?
- jarrett 14y ago
  I would suggest the following informal definition: A patent troll files and seeks to enforce patents of questionable merits, often in large volumes, while engaging in markedly little good-faith research. That's not precise enough to work as a legal standard, but I think it might be a good first step in that direction.
astrodust 14y ago
Is there another way to protect these pure inventors from having their genuine inventions stolen while avoiding the trap of having these non-inventions being used to significantly damage innovation in general? One thing that's missing from the current patent system is having to provide a working implementation. This requirement was removed fairly early in the process since the patent office was being cluttered up with too many scale models or other contraptions. For software patents, one thing that would go a long way would be to make it mandatory and to ensure that the patent describes only the specific innovations expressed in the software and not some hand-wavingly vague potential interpretations. The patent here is so absurd that it's amazing it even got approved. I'm certain that all the techniques described in this document had been done decades prior on many other platforms such as early systems like Desqview, or even the Xerox Alto. The patent clerk must've just rubber-stamped it since it sounded plausible.

JoshTriplett 14y ago

Can you write any description that reliably distinguishes between a patent troll and anyone else who does not use the patent themselves in production (commonly called a "non-practicing entity" (NPE))? If not, that should tell you something. (Not that practicing entities can't resemble patent trolls too, as evidenced by the numerous ongoing smartphone lawsuits.)

karamazov 14y ago
The issue with patent trolls isn't how they use patents - it's the patents that they use. There's nothing wrong with a researcher suing a factory over a patent the researcher holds that the factory is using without license. There's also nothing wrong with the researcher selling the patent to a lawyer, so as to avoid dealing with the legal system, and that lawyer then suing the factory. This works if the patent was issued for a meaningful advance. In fact, the researcher may not have been interested in doing the research (or at least in revealing his methodology) without patent protection. If the methodology were not revealed, other people would need to spend the time to develop their own process, when it would be easier for them to pay the researcher and then use his work. The issue with software patents is that the methodologies covered are trivial, and the time needed to develop your own method is trivial as well. Because the patents are granted for obvious solutions, many people end up violating the patent without ever having seen it; this is why patents are not supposed to be granted for obvious developments.
- AlisdairO 14y ago
  To roughly recall a comment someone made to me on HN a while back, if it's profitable to think of ideas and do nothing with them until you can sue someone else independently coming up with the same idea, there's a problem with the patent system. The whole point of patents is that they're for ideas that it would be unlikely that someone else would come up with on their own.
- Zak 14y ago
  The problem I see with software patents is that they're usually attempts to patent ideas, but ideas aren't patentable. The implementations of most of the ideas in question are obvious and therefore not patentable. I don't think that we, as a society should decide to allow ideas to be owned, so I do think we should crack down on trivial patents.
  - runako 14y ago
    >> we should crack down on trivial patents. I think everyone is in agreement on this, but the definition of "trivial" is the sticking point.
Avenger42 14y ago
What about a distinction between people who license a patent to only one company (e.g., a university researcher creates a fast-and-accurate HIV test, the university licenses it to a company to produce & market it) versus companies who hold a patent and try to collect licensing fees from every company under the sun, most of whom have no connection to another?
jarrett 14y ago
I would suggest the following informal definition: A patent troll files and seeks to enforce patents of questionable merits, often in large volumes, while engaging in markedly little good-faith research. That's not precise enough to work as a legal standard, but I think it might be a good first step in that direction.

astrodust 14y ago

Is there another way to protect these pure inventors from having their genuine inventions stolen while avoiding the trap of having these non-inventions being used to significantly damage innovation in general? One thing that's missing from the current patent system is having to provide a working implementation. This requirement was removed fairly early in the process since the patent office was being cluttered up with too many scale models or other contraptions. For software patents, one thing that would go a long way would be to make it mandatory and to ensure that the patent describes only the specific innovations expressed in the software and not some hand-wavingly vague potential interpretations. The patent here is so absurd that it's amazing it even got approved. I'm certain that all the techniques described in this document had been done decades prior on many other platforms such as early systems like Desqview, or even the Xerox Alto. The patent clerk must've just rubber-stamped it since it sounded plausible.