Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Is software only Mathematics and therefore not patentable ?

  1. Sep 28, 2009 #1
    Is software "only Mathematics" and therefore not patentable ?

    There are arguments in some circlues, that the programs in computers are simply mathematical algorithms, and as such (being "only" math) they should _NOT_ be subject to patents.

    Thier argument is "you cant patent an equation" and software is only "math" (computers after all "compute".

    My counter argument, is that if you want to call software just maths, then you should call everything "just math" I argue that everything can be broken down into equations.

    I have trouble making a clear argument, but I try to say that Math is a scientific endevour and programming or software is the result of engineering, engineering is the application of math, experience, and knowledge and intelligence and working out how to resolve the engineering problem.

    Math, is a scientific endevour, and programming is the application of science, and math.

    I try to say math is a pure and universal language, that you could use as a medium of communications with alians or someone of a different language.

    This would not be possible with a computer program, as it requires an intimate knowledge of the states and functions of the target machine.

    What's your opinion ?

    How would you develop a "case" that software is math, or how would you form a case that software is NOT "just math"?
     
  2. jcsd
  3. Sep 28, 2009 #2

    D H

    User Avatar
    Staff Emeritus
    Science Advisor

    Re: Is software "only Mathematics" and therefore not patentable ?

    I might be wrong, but this smells a bit too much like homework. We have rules against doing your homework for you.

    Have you researched the issue? Do you know whether the patent office currently gives patents for algorithms? For software? Do you know what LZW means?
     
  4. Sep 28, 2009 #3
    Re: Is software "only Mathematics" and therefore not patentable ?

    No it's certainly not homework, long gone are my days of homework and assignments. My background is electronics engineering, and communications and computers. Military trained.

    No, the argument is going around some circles of the FOSS (Free Open Source software) people, who are seeking patent reform for software patents. One of their main arguments is that they consider software "only math" and as such equations cannot be patented.

    I took it the same way as you, I would expect there are specific equations that are patented, if that is the case they have no argument.

    Or if there are not equations that are subject to patents, why would software be exempt from patents due to it only being "simply math".

    I dont agree, that software is just math and cannot be patented. I argue that software is the application of math, just as much as a highrise building design is the application of math.

    I also try to argue that software cannot do all math, but math can do all software, the lowest common factor (to me) is everything is math at some point, But software (computer programs) are not "just math.

    Purely for self interest. not homework or assignment.
     
  5. Sep 28, 2009 #4
    Re: Is software "only Mathematics" and therefore not patentable ?

    yes, I do know that LZW is a compression algorithm, and I do know there are many patents on software. I do not know if there are any patented equations, but there may well be.

    I also know that at some level, a computer does "manipulations" on data, but are those manipulations "just math" and as the Free Software people argue, "as they are only mathematical manipulations" that software "IS" math.

    Im saying math is a scientific endevour, and that engineering is the application of math and knowledge to solve a problem.

    and that software is *NOT* math, but an engineering disipline the results of which if innovative can be patented, just like a new electric motor that too is based on "equations" for it's operation.
     
  6. Sep 28, 2009 #5

    rcgldr

    User Avatar
    Homework Helper

    Re: Is software "only Mathematics" and therefore not patentable ?

    USA patent office started to allow software patents about 20 to 30 years ago. Their record hasn't been that great though, as I'm aware of some software patents that should fail the "obvious" test. My version of this test is give 10 engineers in a specific field a problem to solve, if 7 or more of them come up with the same answer as the proposed software algorithm, then it shoud be considered "obvious".
     
  7. Sep 28, 2009 #6

    D H

    User Avatar
    Staff Emeritus
    Science Advisor

    Re: Is software "only Mathematics" and therefore not patentable ?

    To Jeff: Seven out of ten? I would not call that obvious. I would not even call that blatantly obvious. I would call that so effing obvious that "a caveman could do it". I would say that "novel" means that every one of your ten programmers either failed to come up with anything at all or came up with something that does not bear any resemblance to the proposed algorithm.


    To Darryl: What is with all this "simply math" and "just math" crap? Suppose persons A and B work for months or years on some apparently intractable mathematical (person A) problem / engineering (person B) problem. Both have a series of brilliant aha! moments that lead them to the solutions. Suppose furthermore that these novel solutions have applications that are potentially worth a lot of money. Finally, suppose that if the idea were let loose on the world that anyone with the necessary expertise could see that these are indeed solutions to these problems. Without some kind of protection, all those months/years of intellectual labor become worthless. What, conceptually, is the difference between these situations?
     
  8. Sep 28, 2009 #7
    Re: Is software "only Mathematics" and therefore not patentable ?

    I tend not to give too much credit to obviousness, and yes im aware there are many many patents allready for software, and im aware some are questinable.

    But as for obviousness, it's easy (too easy) to say somethnig is obvious after it's been though of by someone else.

    Everyone knows what an iceskate is, it took some invention to see that if you replaced the blade with wheels on each corner, you could do the same thing without the ice.

    You would think mearly changing the configuration of the wheels from each corner to an "in-line" configuration would be "obvious". Mabey it is after you've seen on.

    But mounting wheels on shoes, IN A LINE, is a seperate patentable (and defendable) patent.

    Im sure a "bubble sort" or a "binary spanning tree" would appear quite obviou as well after you study and understand it.

    But more to the point, an idea based on scientific concepts, (math, physics, electronics etc) IS patentable, even though they are in fact "just math, just physics and so on".

    Again, I believe that engineering innovation in any engineering field, that is new and innovative should if eligable be patented.

    Im on the side of believing that patents are an effective method of promoting and protecting someone's genius.

    I also believe that killing patents on an engineering disipline like software would limit advancement in the field, not promote it.
    (but thats getting a bit political).

    Im just more interested in the opinion of the scientific community, who dont have vested interest in software, but who has strong knowledge of the difference between scientific study and software engineering. (or any science as opposed to science investigation).
     
  9. Sep 28, 2009 #8
    Re: Is software "only Mathematics" and therefore not patentable ?

    I dont like the "just math" terms either D H, and I agree if you work hard and have some "aha" moments, they you apply your "invention" (mabey an equation) to a practical application that solves a problem. Then that practical application is the result of engineering and invention and should be patented.

    It's not the equation you developed that you will patent, it's the practical product of your inventiveness.

    Exactly what I am saying about software, you can come up with an amazing algorithm to do something inventive based on your skills, knowledge and intellect (and math knowledge).

    You invented a practical device, it might be based on your newly derived equation but you patent the method and device.

    I know im not very clear, but I do see your point, my point is it is ** NOT ** "JUST" math, my argument is everything is math, which makes it hard to say that software is NOT math, but my real meaning is sure software is math, but so is everything else.

    Engineering and invention is the application of knowledge, science, logic, definition of a problem and the practical method of resolving that problem.

    Math is the scientific study of numbers, sets and so on. IMHO.
     
  10. Sep 29, 2009 #9

    rcgldr

    User Avatar
    Homework Helper

    Re: Is software "only Mathematics" and therefore not patentable ?

    What I meant by that is there are existing software patents that would fail the 7 out of 10 criteria. Some would fail the 9 out of 10 criteria. Many would fail the 3 or more out of 10.
     
  11. Oct 2, 2009 #10

    chiro

    User Avatar
    Science Advisor

    Re: Is software "only Mathematics" and therefore not patentable ?

    Algorithms per se can not be patented. What can be patented is an invention or a process that has some form of economic endeavour. Basically if a process or invention is novel, has economic value then it can be patented.

    There are of course a lot of conditions for something to be patented. For one the novelty value is defined whereby the invention is nonobvious for someone who is skilled in the art, or in other words if you came up with a compression algorithm, then that algorithm would be novel and not obvious to come up with for another data compression expert. Because of the law (and unlike mathematics) it is subjective but you will that these issues have gone through the Supreme and High courts and as such the definitions for novelty and newness become clearer as these courts end up defining these terms in a clearer manner.
     
  12. Oct 5, 2009 #11
    Re: Is software "only Mathematics" and therefore not patentable ?

    A better point is to look at any digital product, music, movies, games. In the end, they are stored on one or more platforms (such as one or more DVD's) as binary. Being in binary, it is but one really big number. Should you then be able to patent numbers?

    Or even a more interesting question, should some numbers be illegal? Clearly there are illegal images (lets not go into what they are though), and those images can be put onto a computer. In doing so, they become one (really big) number. Should being in possession of that number be a crime?

    I think the answer to this lies in the point of laws. Are the laws there to restrict actual actions, or to attempt to restrict intentions? We should never make a number illegal or patent it, but we should not allow for some to steal others works claiming it is just a number. Thus, the laws should, in theory, attempt to (though this is not possible) measure why one is using this number or that number or this bit of math or that bit of math.


    As to patents, the end idea is that we do not want to allow one to steal from another their efforts, yet at the same time, we do not want to restrict the flow of ideas. If we lean too far towards freedom, there could possibly result very little monetary reward for those who spend time discovering innovations, and since our world requires money to get buy, people would cease to innovate in that respect. The reverse is if we restrict it too much, then things would become bogged down in making sure you are not reinventing the wheel, not to avoid doing extra work, but to avoid a legal issues, coupled with needing to get permission, if not outright purchase, certain 'pieces' of code from others.
     
    Last edited: Oct 5, 2009
  13. Oct 7, 2009 #12
    Re: Is software "only Mathematics" and therefore not patentable ?

    This is what I think:

    Saying that software is only equations is missing the big picture.

    I think there's a reason why legally not all software is freeware. It's more than just equations, there's also the end product. As you said, everything breaks down to equations. The blender and stove in my kitchen had equations put into them. Does that mean I can make an "exact duplicate" and make money off of it? Software has more than equations, but also the big picture of what it is; which is an idea in itself. If there's a software program for a special game idea, why should others steal that idea plus the exact programming and rip the person with a patent off? When you buy a CD with software, you're not actually buying the software itself, but rather a license to use it for one (or three depending on the license); otherwise you'd be able to make copies and sell it.

    I don't know why it makes sense that someone who spends hours on end has to be cheated because someone makes a copy of what he makes and sells it on the market, making the money instead?
     
    Last edited: Oct 7, 2009
  14. Nov 8, 2009 #13

    SixNein

    User Avatar
    Gold Member

    Re: Is software "only Mathematics" and therefore not patentable ?

    Yes, software is just mathematics and nothing more. The only difference between software and mathematics is the symbolic representation used to describe the ideas. As far as computers are concerned, they do not know the difference between data and code. All of the information must be kept separated by the programmer. A computer is nothing more than a beefed up calculator. When a program is written, it is converted into a serious of numbers. Some of these numbers are data, and some of these numbers are code. The screen you are currently viewing is nothing more than a fancy graph.

    For a comparison to electronics, imagine people being allowed to patent ohms law or Maxwell equations.

    Software has two main problems:
    1. Theory is making its way into the patent system.
    2. Most software patents are of extremely poor quality.
    3. Software patents last 20 years. (imagine software of 20 years ago.)
    #1 is the largest problem.
     
  15. Nov 8, 2009 #14

    SixNein

    User Avatar
    Gold Member

  16. Nov 8, 2009 #15

    Moonbear

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Re: Is software "only Mathematics" and therefore not patentable ?

    You CAN patent a novel method for doing something, even if it's "just math." So, I think software would fall into that sort of category, where if the method used by the software is novel, then it is patentable. Otherwise, if you're using the same underlying method and just changing the looks of things, then I think it's more of a copyright protection for the pretty way you arranged the pixels on the screen different from how someone else arranged them.

    As for the term non-obvious, I don't know if that has the ordinary meaning, or if it means something that is not already patented and not freely available in the public domain. One thing I've been learning from my boyfriend is that the world of patent law has its very own dictionary that is quite different from the way any sane person would use the terms they use. And then there are times when they get to define a word any way they want as long as they define it clearly in a patent application. When my boyfriend first started practicing and would get frustrated with the patent examiners, we would have long debates when he'd call me up and ask me random questions like, "What do you think the word 'edge' means? And how about the word, 'end'? Do you think they are the same thing or different things?" Unfortunately, I too often ended up siding with the patent examiner, and started learning to first ask, "And how are YOU defining it today? Is there some context I should know first?" :biggrin: If I remember, I'll ask him what non-obvious means. He'll probably tell me it's obvious. :rolleyes:
     
  17. Nov 8, 2009 #16

    turbo

    User Avatar
    Gold Member

    Re: Is software "only Mathematics" and therefore not patentable ?

    If you have ever written software, you already know that there is a great deal of creativity involved in getting it to work properly while consuming as few system resources as possible. (This seems not to be the MS model, but I digress.) Such innovations are protectable through copyright and patent. Long, long ago, when 386s were servers in small businesses, I wrote an entire inventory-control and point-of-sale program for a manufacturer of wood-harvesting equipment, and delivered it on a floppy. Today, such an application would arrive on a suite of CDs or a DVD or two. My programming was proprietary, and I "sold" it to the businesses for the cost of my coding time and installation, and charged them for ongoing maintenance and upgrades. Custom software on the cheap. If I had been writing general-purpose accounting software, I certainly would have protected my intellectual property and would have compiled all the code to make it very hard to reverse-engineer. "Just math" is not a viable reason to deny intellectual property protection.
     
  18. Nov 9, 2009 #17

    SixNein

    User Avatar
    Gold Member

    Re: Is software "only Mathematics" and therefore not patentable ?

    Creativity? You mean mathematics.

    I don't think anyone is arguing against intellectual property protection, but specifically patent protection. Software is already fully and totally protected by copyright law, and it has trademark protection as well.

    Did you do a patent search on your software to see how many patents you violated? Or did you just release the program and hope it doesn't get noticed by patent owners?

    It's virtually impossible to avoid patent infringement now days.
     
  19. Nov 9, 2009 #18

    SixNein

    User Avatar
    Gold Member

    Re: Is software "only Mathematics" and therefore not patentable ?

    I would also point out that most software patents are being awarded to OTHER industries and not the software industry (Bessen and Meurer 22).
     
  20. Nov 9, 2009 #19

    turbo

    User Avatar
    Gold Member

    Re: Is software "only Mathematics" and therefore not patentable ?

    I mean creativity. When I was tweaking a custom accounting program for a large trucking company (trying to reduce the age of their receivables so that they could get a larger line of credit), I heard the the VP (owner's daughter) complaining to her head dispatcher that reimbursements for fuel taxes had fallen off. Some states reimburse pro-rated amounts for fuel taxes paid in state vs fuel consumed in state. I asked her if she wanted an application that would tell drivers on their various routes (all over the US) where and when to fuel up, when to get a bit of fuel and when to really tank up to minimize fuel-tax losses. It took me about a week to design and write that application, and it was an instant success when applied. The drivers hated me (loss of free lunches and free thermos fill-ups were biggies), the dispatchers liked being able to point to me as the bad guy, and the VP got to be a hero with her father, who implemented the fuel-tax program company-wide. Identify a problem, design a solution, and implement it. "Just math?" That's a pretty naive attitude.
     
  21. Nov 9, 2009 #20

    SixNein

    User Avatar
    Gold Member

    Re: Is software "only Mathematics" and therefore not patentable ?

    If you change the symbols around, you have mathematics. I think you misunderstand just how much creativity is involved in mathematics.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Is software only Mathematics and therefore not patentable ?
  1. Is website a software? (Replies: 4)

Loading...