• Wes

    It would be nice if there was a “natural” way of determining the obviousness of a patent. Even though it would be too radical a change to be accepted in practice, one idea would be to add a “obviousness testing period” lasting about 1/5 of the life of a patent.

    For example, someone could apply for a 10 year patent on “a lead-acid battery design that achieves 50% greater energy density than existing designs”. There would then be a period of 2 years (1/5 of 10 years) where the accomplishments of the patent but not the methods/designs would be made public. If anyone could independently duplicate the publicly stated accomplishments of the patent in the obviousness testing period then the patent would be invalidated.

    In this example, if anyone independently developed a lead-acid battery with energy density as good or better than that claimed by the patent in question in the 2 year obviousness testing period then this patent would be invalidated. If, however, the battery design was truly innovative then the patent would be awarded.

    If someone ever got a very long and very broad patent then it could stifle innovation (for example, a 100 year patent on a 75% more effective treatment for cancer) but, on the other hand, such a a patent would be very difficult to get (in the 20 year – 1/5 of 100 years – obviousness testing period it is very likely that someone else would come up with a 75% more effective treatment for cancer).

    There would also be the concern that very specific patents would be difficult to get because the publicly stated accomplishment of the patent would provide clues as to how to duplicate the patent’s methods. On the other hand, if someone needed to make their patent’s publicly stated accomplishments very specific then it would be a good indication that there were other ways of doing the same thing and that the invention wasn’t really that innovative, anyway.

  • http://zgp.org/~dmarti/ Don Marti

    What if the defendant in an infringement case got a graduate from the middle of the class at a mid-range engineering school, locked him or her in a room with only materials published before the date of the patent application, and gave him or her one working day to solve the problem — and one of the solutions this “person of ordinary skill” came up was the patented technique?

  • http://ozlabs.org/~rusty Rusty Russell

    Finally! The erosion of obviousness is often overlooked in favor of prior-art issues, yet it causes at least as many problems (including patent thickets). For places in the world which don’t have the 1-year grace period, an obviousness test is fairly straightforward.

    1) The examiner and the applicant split the patent application into “problem” and “solution”, where the problem statement includes a list of weaknesses of existing solutions. If you can’t split into “problem” and “solution”, your invention is not “useful”.

    2) Publish the problem, wait a week.

    3) Fold any feedback from step (2) into the prior art search.

    Naturally, there will be attempts to game the system, like all patent procedures. There should be some control on who can submit solutions (a “Friends of the PTO” program?) and they should get a significant share of the filing fee if their solution is used to defeat a claim. Also, solutions must be specific (not, “Test all combinations of the following compounds…”).

    The principle is that (1) you don’t get a patent for finding a non-obvious problem, but for finding a solution, and (2) you don’t deserve a 20+-year monopoly for less than a week’s work.

    For unenlightened countries which have decided patents should be easier to obtain and have the one-year grace period, such a solution would be voluntary. Make those who don’t use it more susceptible to other patent exceptions (introduce a post-grant obviousness test, or even have a patent exception for independent invention if this method is not used).

    Cheers,
    Rusty.

  • Christian

    Question is if people like PFF really wants a working system or if they would be happy enough with one that is broken in a less obvious fashion. Noticed that a lot of people who have defended patents in the past are now out demanding reform. Question is if its because they have actually understood the weaknesses of the patent system, especially software patents, or if they still don’t see the true problems and just want some duct tapte applied in order to make the system cause less grievances and thus, as PFF mention, hinder the opposition against the patent system from continuing to grow.

    The sad fact is that the complete brokenness of the patent system today is causing a lot of people to start thinking about it, which in turn means many of them also come to the realisation that patents in a lot of areas cause much more damage than they help promote advances. If the system gets slightly reformed to not so obviously grieve people the issue could end up falling from the public light and we get stuck with a broken system for a long time.