Forgot your password?
typodupeerror
Math Canada Patents

Canada Courts, Patent Office Warns Against Trying To Patent Mathematics 215

Posted by samzenpus
from the free-numbers dept.
davecb writes "The Canadian Intellectual Property Office (CIPO) has recently published two notices for patent examiners relating to patent interpretation, and in particular computer-related/business method type patents saying: 'for example, what appears on its face to be a claim for an "art" or a "process" may, on a proper construction, be a claim for a mathematical formula and therefore not patentable subject matter.'"
This discussion has been archived. No new comments can be posted.

Canada Courts, Patent Office Warns Against Trying To Patent Mathematics

Comments Filter:
  • ...translators and software won't be patentable!

    • by msauve (701917) on Thursday May 16, 2013 @04:44AM (#43739005)
      Isn't code already math? Any algorithm can be implemented on a Turing machine, which is a mathematical construct.
      • by metageek (466836) on Thursday May 16, 2013 @04:46AM (#43739007)

        Sure it is, but judges still need to be convinced

      • by bunratty (545641)
        Any physical machine can be simulated on a computer, which includes a Turing machine. That doesn't mean the idea is math. In other words, the fact that X can be implemented on a computer does not imply that X is mathematics.
        • by msauve (701917)
          "Simulated," yes, as in imitating, not duplicating. You can simulate a wheel, but that doesn't mean you've violated the wheel patent. And, contrary to what's implied by your claim, that simulation is math, even if the wheel itself isn't. Conversely, taking something physical and simulating it on a computer shouldn't make it subject to patent (the common "x, but on a computer" patents).
      • Any algorithm can be implemented on a Turing machine, which is a mathematical construct.

        You might want to implement it in lambda calculus instead, it's more abstract and probably feels more math-y, even to the judges.

        • by mark-t (151149)
          But why would a person who wants to get his algorithm patented ever want to do that?
      • Isn't code already math? Any algorithm can be implemented on a Turing machine, which is a mathematical construct.

        Any invention is made up of a large number of atoms, which interact with each other according to the laws of quantum physics. Since quantum physics is not patentable, nothing is patentable. Right?

        And actually, not every algorithm _can_ be implemented on a Turing machine for two reasons: 1. A Turing machine running a program can only either stop in a zero state, stop in a one state, or run forever. That's not really very useful. 2. A Turing machine can not be implemented for most algorithms because the Tu

    • by flyneye (84093) on Thursday May 16, 2013 @06:39AM (#43739365) Homepage

      Good, they shouldn't be.
      Many things patentable, shouldn't be, many things unpatentable shouldn't be, but are.
      Patents should expire 4 years after acceptance to promote innovation. If you haven't dug gold out of it in 4 years, it's time to shit or get off the pot. There's a world out there who can innovate. What have you done for us today?

    • by macraig (621737)

      The 19th century called and wants its algorithm [wikipedia.org] back.

    • Careful. Back when I was in college, NO computer programs could be patented under exactly that logic. The whole argument about doing the same as some old business process "on a computer" would have made it NOT patentable, rather than "new" as is currently done. So when the group I worked with came up with a major new wrinkle in a manufacturing process that involved mostly software, the lawyers decided it was unpatentable and we didn't get any filing bonus - unlike the mechanical folks who got a patent an
  • by Esben (553245) on Thursday May 16, 2013 @04:32AM (#43738971)
    but you can patent the application in a specific area to solve a specific problem. Nothing new there.
    • That sounds strange. So you could patent "subtracting two numbers to calculate at what time you have to leave home to arrive on time"?

      • by Anonymous Coward on Thursday May 16, 2013 @05:16AM (#43739095)

        You have to add "with an iphone" to the end, otherwise you could do that with paper and pen and that's not patentable.

      • by tmosley (996283)
        You can patent the process that makes the calculation, not the calculation itself. But IANAPA (patent attorney).
      • You can patent "remember what your regular customers buy and suggest accompaniments and accessories" - which people have been doing since camel caravans on the Silk Road and before - by adding "on a computer" or "with a single click". The concept has gone overboard.
    • by gomiam (587421)
      I think you are wrong. In Europe, for example, a software algorithm is only patentable as part of a physical method and only as far as that physical method is concerned. So you can patent an algorithm that imitates a servomechanism to keep a motor from stalling but that patent only applies to the physical system it is embedded in. I can freely use the same algorithm to, say, control the pressure of a boiler, if it is appliable.

      That is, unless I have misread or forgot something.

  • My guaranteed payday based on my patented use of the letter 'x' in equations suddenly appears in doubt.

    • Re: (Score:3, Funny)

      by Nbrevu (2848029)
      Of course! This is because of prior art. Your so-called 'x' is nothing but a 45 rotation of MY patented '+' symbol for addition.
  • It's not over yet! (Score:5, Informative)

    by bogaboga (793279) on Thursday May 16, 2013 @05:07AM (#43739073)

    Here's why: Lawyers being what they are, will bicker over what exactly a mathematical formula is.

    I will never forget an incident where in the recent Oracle vs Google case, Oracle's side tried to change the facts about a memory reference being symbolic or otherwise. Mind you, this was an expert! It was pathetic!

    Subsequently, the court shot down Oracle's position with this piece.

    The foregoing is sufficient but it is worth adding that Oracle's infringement case was presented through Dr. Mitchell. A reasonable jury could have found his many "mistakes" in his report merely to be convenient alterations to fix truthful admissions earlier made before he realized the import of his admissions. For this reason, a reasonable jury could have rejected every word of his testimony.

    Oracle lost the case - For now.

  • should genes be patentable, but it seems that we maybe have lost that one (sadly)
    • Re:not unlike .. (Score:4, Insightful)

      by Simon Brooke (45012) <stillyet@googlemail.com> on Thursday May 16, 2013 @06:34AM (#43739347) Homepage Journal

      should genes be patentable, but it seems that we maybe have lost that one (sadly)

      Actually, it's the same issue, since genes are information structures which are processed by the cell. Consequently, genes are software, and consequently are mathematics. The fact that we don't yet understand in detail the mechanisms by which the cell processes the information structures is irrelevant.

      If you can't patent software because it is mathematics, then you can't patent genes because they are software.

      • If you can't patent software because it is mathematics, then you can't patent genes because they are software.

        But one of the things that the law defines as patentable is a novel "configuration of matter". A truly original gene would easily qualify under this condition. While the gene is only useful because it is software for a biological system, if it represents a novel configuration of matter to encode that software, then it should be patentable.

        I'm less sure when the case involves taking genes found in nature, and splicing them onto other plants (such as the "Roundup Ready" gene). In this case, while there is

  • by StripedCow (776465) on Thursday May 16, 2013 @05:16AM (#43739093)

    In a little while, when physicists have figured out the laws of the universe, everything might be mathematics... (not just symbolic stuff like computer programs)

    • Re: (Score:2, Funny)

      by L4t3r4lu5 (1216702)
      Ah, the good ol' Slashdot obligatory xkcd [xkcd.com]
    • Re:Matter of time (Score:5, Interesting)

      by khakipuce (625944) on Thursday May 16, 2013 @06:20AM (#43739283) Homepage Journal

      It's the other way round. Mathematics is just an abstract representation of the real world. No amount of physics, maths or theories of "everything" will cure cancer or invent the next IPhone. Patents are about (or at least should be about) the inventive step - i.e. the coming together of several elements to create something new.

      • No amount of physics, maths or theories of "everything" will cure cancer

        Well, a sheet of paper on which a patent is written will not solve many practical problems either.

      • And that seems to apply to software as well. Like many things, software is just a medium for designs and ideas, implemented in math at some levels (usually abstracted away from the designer).

        Not all that different than hardware... though the implications might be.
      • How is teleportation not merely physics? And if you can select and teleport cancer cells, you can cure cancer.

    • physicists have figured out the laws of the universe, everything might be mathematics

      I'm afraid everywhere is already mathematics... This is the difference between physics and mathematics: change the way our Universe works and physics vanish, mathematics (on which physics formula are based on) remain.

  • It really is simple. Software is mathematics because all software can be expressed as mathematical formulas, therefore it can't be patented.

    • Software may be mathematics but the research, engineering and creative thinking that went into bringing the software from concept to actual code isn't.

      If I invented a new artificial intelligence algorithm that has distinctive advantages over every currently known alternative, I believe the principles behind it would still be very much patent-worthy. The exact software implementation is pointless since others would likely be able to rewrite their own alternate implementation from principles and my own implem

      • The purpose of patents is to benefit society same with copyrite (spelling intentional) - to promote a small amount of legitimately deserving works we have to screw up the other 99% and ruin the industry? Free speech includes horrible speech, its the price one pays for the bad minority (and sometimes the bad stuff IS the most important... just as some of the most deserving patents may be the most important.)

        We can no longer afford to cripple research, small business, and even the nation by this idiotic pat

    • It really is simple. Software is mathematics because all software can be expressed as mathematical formulas, therefore it can't be patented.

      Here's a bit of software:
      printf ("Hello, world\n");

      Now you write the formula. Let's see how simple it is.

  • by Stolpskott (2422670) on Thursday May 16, 2013 @06:15AM (#43739257)

    As a clarification and reminder for the patent examiners, this is a good thing. However, the USPTO has guidelines and rules as well, with odd little things like "Prior art" and descriptions of things that should not be patentable.
    However, there is also a policy (not sure if it is written, or just written about) that if the patent examiner cannot understand the patent application but cannot specifically see that it definitely contravenes any of the guidelines for things that should not be patentable, the patent should be granted and then the court system should be used to test the validity of the patent.

    • he examiners are supposed to be at least familiar with the area. of ordinary skill in the art as it were.

      if they cannot understand a patent then how could they build what it describes. so if they cannot understand it then it's not meeting the goal of patents: ie disclosing how something works.

  • Well, I HATE this software argument about patents as, to be honest, EVERYTHING can be described as mathematics. From mechanical systems to genetic code, from electrical designs to source code. You can name anything: I can write it down either as mathematical model using a set of formulas or using a array of numbers. If mechanical designs and electronic systems can be patented so can be software.
    The problem with the current patent system, in particular in the US, is that it is a lousy version of an idea fro
    • Well, I HATE this software argument about patents as, to be honest, EVERYTHING can be described as mathematics.

      It's really just an argument used by lazy thinkers. They don't want software patents (with good reasons, because there is a huge number of software patents that are in my opinion obvious), they read that maths is not patentable, so they shout "software is maths, so it is not patentable". There are a few huge problems with it:

      1. You can of course claim that all software is maths, and call everyone stupid who doesn't believe it, but that doesn't make it true. If someone says "it is a mathematical function"

  • I'm against patients in general, except for big billion dollar medical advancements I don't really see the point. Why can't we just share knowledge, intellect and wisdom and not slap each other with law suits.
  • 2 + 2 = 5 ©

    Nobody says the IP has to be correct.

  • But text can be copyrighted.

    The first person to print out the full value of Pi can copyright it....

"Ahead warp factor 1" - Captain Kirk

Working...