5 ms·

Wait, so are Software patents acceptable but patents on mathematical algorithms not okay? Where is the rational in this? Granted, the patent in question is absu

by jostmey 14y ago

Wait, so are Software patents acceptable but patents on mathematical algorithms not okay? Where is the rational in this? Granted, the patent in question is absurd to a new extreme (trying to patent floating point numbers).

claudius 14y ago
I don’t quite get it either – though I am generally not opposed to software patents on complex software, nor for ‘maths patents’ on complex algorithms. It is just the trivial ones that get annoying over time… On the other hand, most ‘engineering patents’ were probably also quite trivial during the first few years of the modern patent system – but you don’t necessarily have to repeat mistakes :)
- jhaywood 14y ago
  From what I learned of math, all mathematical algorithms are eventually refactored until they are "trivial". Trivial is a mathematician's favorite word.
  - Tyr42 14y ago
    I think I prefer, "by definition", but it's close.
greyfade 14y ago
Mathematics are reasoned to be "a law of nature," and "abstract ideas," and so they are explicitly excluded as unpatentable subject matter under 35 USC 101. Algorithms are established to be mathematics under Case Law, particularly Chakrabarty, 447 U.S. at 309, 206 USPQ at 197; and Flook, 437 U.S. at 585, 198 USPQ at 195. Diehr, 450 U.S. at 188-178; establishes that if the patent is for a particular application of an abstract idea, but not for the abstract idea itself, it may be patentable.
- jlgreco 14y ago
  > Diehr, 450 U.S. at 188-178; establishes that if the patent is for a particular application of an abstract idea, but not for the abstract idea itself, it may be patentable. So what does that mean, "a^2 + b^2 = c^2" cannot be patented, but "using 'a^2 + b^2 = c^2' to find the length^2 of the hypotenuse of a right triangle" in theory could be (were it "invented" today of course)?
  - greyfade 14y ago
    If it were being used to create a measuring implement, perhaps. But yes, you've got the gist of it.
  - greyfade 14y ago
    Actually, I should clarify a little: The more recent In re Bilski establishes that a rote process (like finding the length of the hypotenuse in your example) is not patentable. It requires a machine or a tangible transformation. That is, a machine that does this calculation.
- AnthonyMouse 14y ago
  >Diehr, 450 U.S. at 188-178; establishes that if the patent is for a particular application of an abstract idea, but not for the abstract idea itself, it may be patentable. Worth pointing out is that the patent in Diehr claimed a method for curing rubber and contemplated a collection of specialized hardware. Not exactly pure software. Do you know of any case other than Diehr where the Supreme Court has actually sanctioned a software patent?
  - greyfade 14y ago
    Not personally, no. Other than Diehr and Flook, SCOTUS hasn't really dealt with software patents much at all. One of the dissenting justices in Bilski lamented the fact that they didn't deal with the wider issue of system and process patents, particularly as they applied to software.
pbhjpbhj 14y ago
>the patent in question is absurd to a new extreme // Not really new nor an extreme. In the past someone claimed they had a patent on arabic numerals that covered use of them in representation of numbers ... the application was refused ...
tzs 14y ago
> Wait, so are Software patents acceptable but patents on mathematical algorithms not okay? Where is the rational in this? Compare to other fields. Patents on drugs are allowed, for example, but patents on the laws of chemistry are not. Patents on mechanical devices are allowed, but patents on physical laws are not.
- jlgreco 14y ago
  Could you patent the application of a law of physics/chemistry, where a new one discovered? Or could Feynman have patented using Feynman diagrams to visualize interactions between particles?
  - greyfade 14y ago
    In principle, no. They are laws of nature and therefore not subject to patentability.
- betterunix 14y ago
  The difference is that an algorithms patent is a patent on a mathematics; a patent on a drug is not a patent on a law of nature. That is why software patents try to make clever end-runs around the restrictions on patents by saying things like, "...on a computer."

by jostmey 14y ago

claudius 14y ago

I don’t quite get it either – though I am generally not opposed to software patents on complex software, nor for ‘maths patents’ on complex algorithms. It is just the trivial ones that get annoying over time… On the other hand, most ‘engineering patents’ were probably also quite trivial during the first few years of the modern patent system – but you don’t necessarily have to repeat mistakes :)

jhaywood 14y ago
From what I learned of math, all mathematical algorithms are eventually refactored until they are "trivial". Trivial is a mathematician's favorite word.
- Tyr42 14y ago
  I think I prefer, "by definition", but it's close.

greyfade 14y ago

Mathematics are reasoned to be "a law of nature," and "abstract ideas," and so they are explicitly excluded as unpatentable subject matter under 35 USC 101. Algorithms are established to be mathematics under Case Law, particularly Chakrabarty, 447 U.S. at 309, 206 USPQ at 197; and Flook, 437 U.S. at 585, 198 USPQ at 195. Diehr, 450 U.S. at 188-178; establishes that if the patent is for a particular application of an abstract idea, but not for the abstract idea itself, it may be patentable.

jlgreco 14y ago
> Diehr, 450 U.S. at 188-178; establishes that if the patent is for a particular application of an abstract idea, but not for the abstract idea itself, it may be patentable. So what does that mean, "a^2 + b^2 = c^2" cannot be patented, but "using 'a^2 + b^2 = c^2' to find the length^2 of the hypotenuse of a right triangle" in theory could be (were it "invented" today of course)?
- greyfade 14y ago
  If it were being used to create a measuring implement, perhaps. But yes, you've got the gist of it.
- greyfade 14y ago
  Actually, I should clarify a little: The more recent In re Bilski establishes that a rote process (like finding the length of the hypotenuse in your example) is not patentable. It requires a machine or a tangible transformation. That is, a machine that does this calculation.
AnthonyMouse 14y ago
>Diehr, 450 U.S. at 188-178; establishes that if the patent is for a particular application of an abstract idea, but not for the abstract idea itself, it may be patentable. Worth pointing out is that the patent in Diehr claimed a method for curing rubber and contemplated a collection of specialized hardware. Not exactly pure software. Do you know of any case other than Diehr where the Supreme Court has actually sanctioned a software patent?
- greyfade 14y ago
  Not personally, no. Other than Diehr and Flook, SCOTUS hasn't really dealt with software patents much at all. One of the dissenting justices in Bilski lamented the fact that they didn't deal with the wider issue of system and process patents, particularly as they applied to software.

pbhjpbhj 14y ago

>the patent in question is absurd to a new extreme // Not really new nor an extreme. In the past someone claimed they had a patent on arabic numerals that covered use of them in representation of numbers ... the application was refused ...

tzs 14y ago

> Wait, so are Software patents acceptable but patents on mathematical algorithms not okay? Where is the rational in this? Compare to other fields. Patents on drugs are allowed, for example, but patents on the laws of chemistry are not. Patents on mechanical devices are allowed, but patents on physical laws are not.

jlgreco 14y ago
Could you patent the application of a law of physics/chemistry, where a new one discovered? Or could Feynman have patented using Feynman diagrams to visualize interactions between particles?
- greyfade 14y ago
  In principle, no. They are laws of nature and therefore not subject to patentability.
betterunix 14y ago
The difference is that an algorithms patent is a patent on a mathematics; a patent on a drug is not a patent on a law of nature. That is why software patents try to make clever end-runs around the restrictions on patents by saying things like, "...on a computer."