6 ms·

1 + 1 (pat. pending) — Mathematics, Software and Free Speech

97 points by wisesage5001 15y ago

SkyMarshal 15y ago
TLDR: There's no way to tldr this, it's a pretty awesome essay. It's possibly the most extensive and elaborate essay I've ever read on an actual web page instead of a PDF or journal. Basically the author says: "This article provides a detailed factual explanation of why software is mathematics, complete with the references in mathematical and computer science literature. It also includes a detailed factual explanation of why mathematics is speech, complete once again with references." The implication being that if software is mathematics (and/or speech), it can't be patented. IANAL but I've suspected that a clear and elaborate delineation of the equivalence of software and math/logic has been one big things missing in this debate, and its absence has allowed patent trolls to prosper much more than they would have otherwise. Does anyone know of any other efforts like this to show software == math == logic?
- corin_ 15y ago
  You began with "there's no way to tldr this" then moved on with "Basically the author says" and sumarised it?
  - chc 15y ago
    The obvious meaning to me is "Here's a summary. But no summary could really do justice to this article. You should read it."
    - SkyMarshal 15y ago
      Exactly.
  - SkyMarshal 15y ago
    Yeah, I was in a hurry as I was writing it and my brain got ahead of my typing and I completely forgot to add this line after my quotation of the author: > Then he goes on to do exactly that throughout the rest of the article. Yes it somewhat contradicts my assertion that there's no way to tldr it, but good tldr's usually condense not just the author's thesis but some of the most pertinent details as well. But in this case that is very difficult, at best.
- brlewis 15y ago
  As I recall software == math == logic was one of the principles behind the Benson and Flook decisions of the U.S. Supreme Court. Then came Diehr, which didn't say anything had changed, but it was opportunistically read as if something had changed. Software is patentable today because our legal system is inconsistent. Law is not law.
  - micks56 15y ago
    In Benson and Flook, the claimed inventions were on algorithms and NOT tied to a machine. In Diehr, the inventor claimed "algorithm + machine" where machine was a computer. The Court looks at a computer as an infinitely configurable machine, with each new software algorithm creating a new invention. That is the distinction that the courts make between Benson/Flook and Diehr, which has allowed "software" patents to exist.
    - brlewis 15y ago
      That's grossly inaccurate. The rest of the Diehr patent included a combination of elements claimed to be novel, not a general-purpose computer.
      - micks56 15y ago
        The novelty was using a computer to do this, which no one had done before. Prior to this invention, curing rubber was done manually. Also, from my understanding an infinitely configurable computer and general-purpose computer aren't the same thing. Infinitely configurable means a new machine every time new software is installed, which would allow the invention to fall within the patent statutes. A general-purpose computer is not a new machine, has already been invented, and therefore trying to patent an "algorithm + GPC" would mean a patent on the algorithm, which is outside patent protection.
        brlewis 15y ago
        Your description of the claimed novelty differs wildly from the majority opinion in Diehr: http://brlewis.com/y/2004/diehr.htm http://brlewis.com/y/2004/diehr.htm According to the respondents, the continuous measuring of the temperature inside the mold cavity, the feeding of this information to a digital computer which constantly recalculates the cure time, and the signaling by the computer to open the press, are all new in the art.
shasta 15y ago
Very long winded article that doesn't really do a good job of laying out a case against software patents. Despite its length, the article doesn't address the core false dichotomy: software is both an encoding of a pure mathematical entity and the input to a machine that actually does stuff. Just beating us over the head with "software is mathematics/speech" for 50 pages doesn't address the real issue which is that it isn't just mathematics/speech.
alain94040 15y ago
I love Groklaw 99% of the time, and this is one case where the article doesn't make any sense to me. Sure, at the most abstract level, software is math and math is speech. But "First Amendment issues resulting from exclusive rights granted to the exercise of mathematical speech"??? The fact that running a piece of software is predictable (the same program runs the same way with the same inputs), is not a strong argument. A machine (with physical gears) will perform the same function when run in the exact same conditions. Just because one is an analog machine and one is a digital machine doesn't change the concept of invention: making something new from existing parts.
- jbri 15y ago
  The predominant claim is that the machine has already been invented, and a new input to the existing machine should not be patentable.
- baddox 15y ago
  Some people (okay, maybe just me) aren't huge fans of intellectual property protection because, like you say, all invention/creation is making something new from existing parts. The property analogy doesn't hold up in my opinion, especially in the Information Age.
- jamii 15y ago
  A much stronger argument against patents, which the OP only touched upon, is the existence of FOSS. Patents exist to encourage the disclosure of ideas but FOSS already provides an excellent mechanism for this. Not only that, but the existence of software patents is damaging to the FOSS community which generally cannot afford to navigate the idea minefield. Software patents don't even require disclosing enough information to actually implement the idea so in many cases there is zero benefit to society. Judges also need to be made aware just how much of the modern information age is built upon FOSS.
Detrus 15y ago
I don't see why it matters if software is math or not. If the goal of the patent system is to encourage innovation and patenting software or math accomplishes it, then patent it. If not, don't patent it. The real difference is in how it's used and created. Generally speaking everyone uses math, so if you could patent it you'd slow innovation for everyone. If you're patenting algorithms that take a few days to create and they can be applied across many domains, you're slowing innovation. Today this applies to math and software. It will soon apply to engineering physical objects as 3D printers, nanotech, etc.. will make it easier and faster. Basically if innovation in a field is easy enough, patents will slow it down and should not apply. Obviously programmers want software to be special but if it was, the patent system would be less consistent than it currently is. If you want software to make sense in the patent system, just redesign the whole thing.
- fryguy 15y ago
  The problem is that math isn't "invented", it's "discovered". Hence why it's traditionally been un-patentable.
  - Detrus 15y ago
    These sound like silly distinctions. I'm sure lawyers have good rhetoric to back themselves up but maintaining the pace of innovation should be prioritized above their greatest arguments. If there was an opportunity to redesign the patent system, some sort of economic framework should be used instead of the legal one.
    - fryguy 15y ago
      Well there is a fundamental difference between the two. The light bulb was invented, was lightning was discovered. It already existed, and was merely discovered. It was discovered that there are are infinite primes. It was discovered that f'(x) = 2x when f(x) = x^2. It was discovered that when p is prime, (a * x) === b mod p, x is unique mod p (unless a=0, or something else). Eventually you get to things like cryptography where you can apply some simple things that were discovered into a process, and come up with encryption. If you go back to mathematics and consider (Newton's method)[http://en.wikipedia.org/wiki/Newtons_method http://en.wikipedia.org/wiki/Newtons_method], that seems like it might have been invented, but it's derived from a series of discoveries, and hence is considered a discovery rather than an invention. If you used the Newton-Rhapson method to calculate the zero of a function in code somewhere, you should feel safe that it's not protected by a patent (not to mention it was discovered ages ago). Since computer science is applied mathematics, and mathematics are discovered rather than invented, it makes sense that it should be protected as well. I'm not the best at explaining things, but hopefully you can understand the difference.
fryguy 15y ago
I don't understand why the author needs to bring FOSS into the discussion. The same argument could be made for regular patents in as much as someone could invent something like a way to make water potable, and then a charity wants to mass-produce them and give them to poor people in Africa, and the patent would prevent that. See pharmaceutical patents for a similar thing. There are other arguments as to why it interacts with FOSS software (patent minefields etc) that affect regular software developers too, so there isn't a need to separate FOSS from regular software development. Those aspects should be the focus, not on the harm to the public.
bfe 15y ago
Ultimately any physical interaction can be abstracted into math - ultimately, in precisely the same way as software being executed by a computing system. I keep coming back to a recognition of every argument against software patents ultimately reducing to an argument against the concept of patents in general.
- azakai 15y ago
  > Ultimately any physical interaction can be abstracted into math - ultimately, in precisely the same way as software being executed by a computing system. I agree on this point. > I keep coming back to a recognition of every argument against software patents ultimately reducing to an argument against the concept of patents in general. But I completely disagree on this one. I think the 'software is math' argument is not convincing against patents (and I have a background in mathematics, so you would think I would like such an argument ;). But there are plenty of other excellent arguments against software patents in particular, that are not valid against other kinds of patents. Patents make sense when progress is fairly slow, people read patents, and patents are granted for actual innovation. Then you do want to award a patent for the rare actual invention - it helps speed progress!. These 3 conditions used to be true, more or less, for patents in general. But today none of them are true for software patents: 1. Progress is ridiculously fast. The entire industry changes in just a few years. And patents last for decades! 2. Software engineers do not read patents. Both because there is no actual benefit to doing so, and because legal counsel always says "do not read patents." 3. A huge amount of patents are granted every year, and the quality is very low. We constantly hear about ridiculous patents being granted and enforced. (And yes, I know that headlines and Slashdot summaries are misleading - you need to read the claims. But even when you do read those claims, in most cases the ridiculous patents are still ridiculous.) So I am against software patents, as they are harmful to the industry in their current form. That is more than enough of a reason, regardless of whether software is math or not (it is). Whereas, patents in general may still be useful in other fields, and my arguments against software patents are not relevant to them.
  - bfe 15y ago
    You've articulated this position about as well as I've seen it put. I'm far from an unbiased party on these issues. I've been thinking a lot on how they should be analyzed. There certainly have been quality issues, and it's vital that patents serve their intended purpose of nurturing and protecting innovation.
zachrose 15y ago
See also, this episode of Dinosaur Comics: http://www.qwantz.com/index.php?comic=353 http://www.qwantz.com/index.php?comic=353 "Any series of 1s and 0s can be converted (quite easily) into a regular number! So, for instance, your favorite song is contained—quite literally—in one single, 4 million digit number."

97 points by wisesage5001 15y ago

TLDR: There's no way to tldr this, it's a pretty awesome essay. It's possibly the most extensive and elaborate essay I've ever read on an actual web page instead of a PDF or journal. Basically the author says: "This article provides a detailed factual explanation of why software is mathematics, complete with the references in mathematical and computer science literature. It also includes a detailed factual explanation of why mathematics is speech, complete once again with references." The implication being that if software is mathematics (and/or speech), it can't be patented. IANAL but I've suspected that a clear and elaborate delineation of the equivalence of software and math/logic has been one big things missing in this debate, and its absence has allowed patent trolls to prosper much more than they would have otherwise. Does anyone know of any other efforts like this to show software == math == logic?

corin_ 15y ago
You began with "there's no way to tldr this" then moved on with "Basically the author says" and sumarised it?
- chc 15y ago
  The obvious meaning to me is "Here's a summary. But no summary could really do justice to this article. You should read it."
  - SkyMarshal 15y ago
    Exactly.
- SkyMarshal 15y ago
  Yeah, I was in a hurry as I was writing it and my brain got ahead of my typing and I completely forgot to add this line after my quotation of the author: > Then he goes on to do exactly that throughout the rest of the article. Yes it somewhat contradicts my assertion that there's no way to tldr it, but good tldr's usually condense not just the author's thesis but some of the most pertinent details as well. But in this case that is very difficult, at best.
brlewis 15y ago
As I recall software == math == logic was one of the principles behind the Benson and Flook decisions of the U.S. Supreme Court. Then came Diehr, which didn't say anything had changed, but it was opportunistically read as if something had changed. Software is patentable today because our legal system is inconsistent. Law is not law.
- micks56 15y ago
  In Benson and Flook, the claimed inventions were on algorithms and NOT tied to a machine. In Diehr, the inventor claimed "algorithm + machine" where machine was a computer. The Court looks at a computer as an infinitely configurable machine, with each new software algorithm creating a new invention. That is the distinction that the courts make between Benson/Flook and Diehr, which has allowed "software" patents to exist.
  - brlewis 15y ago
    That's grossly inaccurate. The rest of the Diehr patent included a combination of elements claimed to be novel, not a general-purpose computer.
    - micks56 15y ago
      The novelty was using a computer to do this, which no one had done before. Prior to this invention, curing rubber was done manually. Also, from my understanding an infinitely configurable computer and general-purpose computer aren't the same thing. Infinitely configurable means a new machine every time new software is installed, which would allow the invention to fall within the patent statutes. A general-purpose computer is not a new machine, has already been invented, and therefore trying to patent an "algorithm + GPC" would mean a patent on the algorithm, which is outside patent protection.
      - brlewis 15y ago
        Your description of the claimed novelty differs wildly from the majority opinion in Diehr: http://brlewis.com/y/2004/diehr.htm http://brlewis.com/y/2004/diehr.htm According to the respondents, the continuous measuring of the temperature inside the mold cavity, the feeding of this information to a digital computer which constantly recalculates the cure time, and the signaling by the computer to open the press, are all new in the art.

shasta 15y ago

Very long winded article that doesn't really do a good job of laying out a case against software patents. Despite its length, the article doesn't address the core false dichotomy: software is both an encoding of a pure mathematical entity and the input to a machine that actually does stuff. Just beating us over the head with "software is mathematics/speech" for 50 pages doesn't address the real issue which is that it isn't just mathematics/speech.

alain94040 15y ago

I love Groklaw 99% of the time, and this is one case where the article doesn't make any sense to me. Sure, at the most abstract level, software is math and math is speech. But "First Amendment issues resulting from exclusive rights granted to the exercise of mathematical speech"??? The fact that running a piece of software is predictable (the same program runs the same way with the same inputs), is not a strong argument. A machine (with physical gears) will perform the same function when run in the exact same conditions. Just because one is an analog machine and one is a digital machine doesn't change the concept of invention: making something new from existing parts.

jbri 15y ago
The predominant claim is that the machine has already been invented, and a new input to the existing machine should not be patentable.
baddox 15y ago
Some people (okay, maybe just me) aren't huge fans of intellectual property protection because, like you say, all invention/creation is making something new from existing parts. The property analogy doesn't hold up in my opinion, especially in the Information Age.
jamii 15y ago
A much stronger argument against patents, which the OP only touched upon, is the existence of FOSS. Patents exist to encourage the disclosure of ideas but FOSS already provides an excellent mechanism for this. Not only that, but the existence of software patents is damaging to the FOSS community which generally cannot afford to navigate the idea minefield. Software patents don't even require disclosing enough information to actually implement the idea so in many cases there is zero benefit to society. Judges also need to be made aware just how much of the modern information age is built upon FOSS.

Detrus 15y ago

I don't see why it matters if software is math or not. If the goal of the patent system is to encourage innovation and patenting software or math accomplishes it, then patent it. If not, don't patent it. The real difference is in how it's used and created. Generally speaking everyone uses math, so if you could patent it you'd slow innovation for everyone. If you're patenting algorithms that take a few days to create and they can be applied across many domains, you're slowing innovation. Today this applies to math and software. It will soon apply to engineering physical objects as 3D printers, nanotech, etc.. will make it easier and faster. Basically if innovation in a field is easy enough, patents will slow it down and should not apply. Obviously programmers want software to be special but if it was, the patent system would be less consistent than it currently is. If you want software to make sense in the patent system, just redesign the whole thing.

fryguy 15y ago
The problem is that math isn't "invented", it's "discovered". Hence why it's traditionally been un-patentable.
- Detrus 15y ago
  These sound like silly distinctions. I'm sure lawyers have good rhetoric to back themselves up but maintaining the pace of innovation should be prioritized above their greatest arguments. If there was an opportunity to redesign the patent system, some sort of economic framework should be used instead of the legal one.
  - fryguy 15y ago
    Well there is a fundamental difference between the two. The light bulb was invented, was lightning was discovered. It already existed, and was merely discovered. It was discovered that there are are infinite primes. It was discovered that f'(x) = 2x when f(x) = x^2. It was discovered that when p is prime, (a * x) === b mod p, x is unique mod p (unless a=0, or something else). Eventually you get to things like cryptography where you can apply some simple things that were discovered into a process, and come up with encryption. If you go back to mathematics and consider (Newton's method)[http://en.wikipedia.org/wiki/Newtons_method http://en.wikipedia.org/wiki/Newtons_method], that seems like it might have been invented, but it's derived from a series of discoveries, and hence is considered a discovery rather than an invention. If you used the Newton-Rhapson method to calculate the zero of a function in code somewhere, you should feel safe that it's not protected by a patent (not to mention it was discovered ages ago). Since computer science is applied mathematics, and mathematics are discovered rather than invented, it makes sense that it should be protected as well. I'm not the best at explaining things, but hopefully you can understand the difference.

fryguy 15y ago

I don't understand why the author needs to bring FOSS into the discussion. The same argument could be made for regular patents in as much as someone could invent something like a way to make water potable, and then a charity wants to mass-produce them and give them to poor people in Africa, and the patent would prevent that. See pharmaceutical patents for a similar thing. There are other arguments as to why it interacts with FOSS software (patent minefields etc) that affect regular software developers too, so there isn't a need to separate FOSS from regular software development. Those aspects should be the focus, not on the harm to the public.

bfe 15y ago

Ultimately any physical interaction can be abstracted into math - ultimately, in precisely the same way as software being executed by a computing system. I keep coming back to a recognition of every argument against software patents ultimately reducing to an argument against the concept of patents in general.

azakai 15y ago
> Ultimately any physical interaction can be abstracted into math - ultimately, in precisely the same way as software being executed by a computing system. I agree on this point. > I keep coming back to a recognition of every argument against software patents ultimately reducing to an argument against the concept of patents in general. But I completely disagree on this one. I think the 'software is math' argument is not convincing against patents (and I have a background in mathematics, so you would think I would like such an argument ;). But there are plenty of other excellent arguments against software patents in particular, that are not valid against other kinds of patents. Patents make sense when progress is fairly slow, people read patents, and patents are granted for actual innovation. Then you do want to award a patent for the rare actual invention - it helps speed progress!. These 3 conditions used to be true, more or less, for patents in general. But today none of them are true for software patents: 1. Progress is ridiculously fast. The entire industry changes in just a few years. And patents last for decades! 2. Software engineers do not read patents. Both because there is no actual benefit to doing so, and because legal counsel always says "do not read patents." 3. A huge amount of patents are granted every year, and the quality is very low. We constantly hear about ridiculous patents being granted and enforced. (And yes, I know that headlines and Slashdot summaries are misleading - you need to read the claims. But even when you do read those claims, in most cases the ridiculous patents are still ridiculous.) So I am against software patents, as they are harmful to the industry in their current form. That is more than enough of a reason, regardless of whether software is math or not (it is). Whereas, patents in general may still be useful in other fields, and my arguments against software patents are not relevant to them.
- bfe 15y ago
  You've articulated this position about as well as I've seen it put. I'm far from an unbiased party on these issues. I've been thinking a lot on how they should be analyzed. There certainly have been quality issues, and it's vital that patents serve their intended purpose of nurturing and protecting innovation.

zachrose 15y ago

See also, this episode of Dinosaur Comics: http://www.qwantz.com/index.php?comic=353 http://www.qwantz.com/index.php?comic=353 "Any series of 1s and 0s can be converted (quite easily) into a regular number! So, for instance, your favorite song is contained—quite literally—in one single, 4 million digit number."