Why we need software patents, part 1 of 3

Why we need software patents, part 1 of 3

Summary: The hardware/software distinction is another of those oft-repeated but intellectually nonsensical distinctions that we need to do away with if we are to have an intelligent discussion about software patents.

TOPICS: Patents

In this week's ZDNet Great Debate, I argued in defense of software patents. This is not a popular position among heavy internet users, and in the popular vote I lost 89% to 11% (I'm relieved to say that the moderator declared me the winner). I thought I'd take a crack at picking up another percent by doing three posts about the basic premises of the argument for software patents.

In this first of three posts, I want to address what I think is the number one fallacy of the anti-software-patents position: the assumption that software and hardware should be treated differently.

Software is not the same as mathematical formulas or literary text. Software is much more like a machine. A typewriter is no more of an invention than word processing software -- I'd actually argue that word processing software requires considerably more inventiveness. If a person comes up with an innovative new idea, the idea should be patentable regardless of whether the implementation is via a traditional mechanical machine or a virtual software machine. Once you understand the nature of software in this way, software patents make perfect sense.

One of the best pieces ever written on this subject comes from Martin Goetz, who among other things was the recipient of the first-ever software patent in 1968. As he puts it:

The choice of implementation for computer functions is a pure economic choice which mainly has to do with cost, speed, and flexibility. Patent applications normally show the preferred implementation and the patent must disclose the invention adequately for one skilled in the art. But the disclosure could be in the form of circuitry for a hardware implementation or a flow chart for a software implementation or a combination of both. Many professionals view software development as building a software machine.

As technology develops and disruptive inventions like the 3D printer become more standard, it will get harder to maintain the hardware/software distinction with a straight face. The hardware/software distinction is another of those oft-repeated but intellectually nonsensical distinctions that we need to do away with if we are to have an intelligent discussion about software patents.

See also:

Topic: Patents

Steven Shaw

About Steven Shaw

Steven Shaw used to be a litigation attorney at Cravath, Swaine &gMoore, a New York law firm, and is now the online community managergfor eGullet.org and the Director of New Media Studies at thegInternational Culinary Center.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Tell us why

    I'm under the impression that lawyers who rely on dogmatic assertions instead of telling judges and jurors why they're right lose.

    The case you really need to make is that software patents really do foster progress in science and the useful arts. I see no evidence of that, rather the reverse. This is made worse by the obscure language of patent applications and the triviality of many of the patents actually granted. In a properly run patent system, it should be nearly impossible to accidentally infringe on a patent, but instead it's all too easy.
    John L. Ries
    • Plagiarism

      As a programmer with more than 30 years experience, I am totally against the USA Software Patent System. My biggest problem is, 90% of the "new" patents I read are industry best practice from the 80s, with the word "mobile" thrown in, to make it a "new and unique" invention.

      Doing it on a pocket based computer, instead of a laptop, desktop or mainframe isn't new or innovative. If I took somebody else's dissertation, put the word mobile on it and handed it in, the University would give me a failing mark for plagiarism, but the US PTO can't seem to get enough of it!

      Searching text and pulling out phone numbers (among other things) and offering to do something with them is something we've been doing for decades. Microsoft hit the big time with it in Office 98, Internet Explorer has done it for donkey's years, yet Apple managed to get it patented a couple of years ago, because their patent application included the word "mobile", at least that is how it seems.

      How is doing the same-old-same-old on a mobile platform "new and innovative" and worthy of a patent?

      And no, I'm not picking on Apple in particular here, Googarola, Microsoft, Nokia, Samsung, htc etc. all try and do the same thing - and that is before we actually get to the patent trolls.

      Companies running round and buying up patents just so they can sue people should be barred from sueing, they were neither innovative, nor are they trying to make money from "their" invention, they are trying to make money out of people who are trying to innovate.

      Add to that the author's statement that software is more like a machine that art or literature; as a programmer, I find that totally bogus. For some processes within an application, there are a minimal number of ways of accomplishing a task, but for many large processes and complete systems, there are myriad ways of actually programming them. That is why I much prefer to see such work covered by copyright only and not by patents.

      And one reason why I am glad I am in Europe and not the USA. At least here, the courts have agreed that software is not patentable.
      • Agree

        Amen. Your position makes a lot more sense than the author's.
  • Clarification... perhaps this will be addressed in parts 2 or 3.

    There are three camps here, not two.

    1. All software patents (and indeed - all patents) are immoral.
    2. Software patents are ok - but they're implemented like classic patents and don't reflect the different market that information technology has.
    3. Everything's just fine the way it is.

    I think at some level most of the 'intense internet users' you comment on are really in the second camp, or would be if a civil discussion happened.

    The chief problem for patents in the information tech market is that the original intention of patents was for processes and inventions with *long* lifetimes. A way to make a better band for a tractor engine is something that's likely to be used for decades if not a century or more. Same with drug patents - a useful drug has a lifespan of decades. The typical patent lifespan (21 years or so) isn't going to kill a market because the technology will still be in demand long after the patent expires.

    But software typically has a lifespan of years and has a vastly wider market than classic patentable ideas have. A good example is the famous LZW patent fight that happened when Compuserve created the GIF file format. Not only did the UniSys patent cause an amazing amount of headaches (because the number of people who were using the patented technology by way of Compuserve's SDK was huge), the patent wasn't asserted until GIF had essentially become a standard.

    In response, PNG and JPG formats were created and the problem was (after a few years and a few big payouts) resolved... but GIF is dead and it's unlikely LZW will ever be used in any graphic file format ever again. The existence of the patent didn't promote its use - it killed it.

    What makes software patents untenable as they are now is because they weren't intended for a market that moves so fast. If they were shortened to say five years, and if the mechanism for validating and issuing them - along with better mechanisms for *finding* them to make sure one isn't in violation - then software patents would be ok.

    What worries a lot of people who seem anti-patent is that they've seen what's happened to copyrights. What started out as a great idea - a limited monopoly that has relatively short existence (30 years) then turns it over to the public domain and excludes private copying or sharing without financial gain from being actionable - has morphed into 70+ years and growing with prohibitions on most forms of copying for personal or non-financial use. As long as Disney owns Mickey Mouse, we can be assured that copyrights will get longer and longer and more and more restrictive. The concern is that software patents will take a similar direction, when they're already too long and onerous.

    There's also a concern that patents can used and seem to be more and more actually used to bludegon the competition. I'm not going to point fingers here - but it seems like rather than using patent law to negotiate reasonble fees, more and more they're used to try and ban products. That's *not* healthy for the economy or for competition.

    In the end, most developers want to make things. A lot of what they do they invent on the fly and don't bother patenting. It's annoying then to suddenly find themselves on the receiving end of a patent lawsuit - especially so late in the process that the very lawsuit essentially kills their company. They just don't have the time to research every algorithm implementation to make sure it's not violating patents. And if that's actually how the IP world thinks it should work, then sorry - they're never going to make allies of most programmer.

    We have better things to do with our time.
    • Fourth party

      Patents on machines, physical objects, and industrial processes are fine, or would be if the system were properly administered, but software and business method patents are not only counterproductive, but a distortion of the law Congress actually enacted and should be rejected by the courts as such.

      That is my position and that of many others.

      Reply to wackoae:

      Thank you for putting words in my mouth. I didn't say there is no software innovation and don't believe that. But there was lots of innovation before there were software patents and no evidence that I know of that would suggest that software patents have fostered greater innovation. In fact, they've made every software developer in the U.S. a potential target of litigation, forcing them to spend more time on legal issues and less on development.

      The purpose of patents is to foster progress in science and the useful arts, not to secure the supposed right of inventors to control their creations. I challenged Mr. Shaw to show how software patents do the former. Care to give him a hand?

      I agree, though, that the bigger problem is with how the USPTO is run and the patents that it actually grants. The system needs to be fixed whether we allow software patents or not.
      John L. Ries
      • Spend your money/time developing some software product ...

        ... that does something nobody else has done before, then watch as a big money corp steals and copies the idea and out-sells you out of business .... then tell us why software should not be patentable.

        The fact that you can't see that there is a lot to innovate in software shows how clueless you really are. For people like you it is OK for others to come out and destroy little company by stealing the ideas and R&D work. In software everything is "easy" after somebody shows you how to do the work.

        Now, the issue of today's patent system are (applicable to more than just software):
        #1- An incompetent Patent Office that doesn't even do a simple search to verify that a submission has prior art.

        #2- Patents with little to no real definitions are approved ... because they are written to look "smart" and the patent reviewer is too incompetent to understand that "dihydrogen oxyde" (H20) is water.

        #3- Patents are approved without a working product. Companies can patent broad ideas taken from anywhere and wait until somebody actually develops the product.

        #4- Patent troll companies are allowed to exist. A patent should only be transferable if the buyer produces a related product.
      • How do they foster progress?

        Patents were a means of giving an inventor time to bring an idea to fruition so that they could make some money off of it, before it became public 'property'.

        It was the opportunity to actually get it done without impediments that was supposed to enable quality implementations to be built upon by others, thereby benefitting all.

        The problem is that companies are expecting to make almost PERPETUAL incomes from patents and copyrights, rather than a HEAD START. It is this attitude that stiffles innovation, not the ideas of patents and copyright of themselves.

        Many years ago the line of the Australian Inventors Association was that rather than patents, it was probably better to make an idea uncopyable for six months to give you enough chance to monetise it. However, litigation and the protracted time to reach global markets makes that option moot.
      • re: Spend your money/time developing some software product ...

        "Spend your money/time developing some software product ...

        ... that does something nobody else has done before, then watch as a big money corp steals and copies the idea and out-sells you out of business .... then tell us why software should not be patentable."

        Except that it doesn't happen that way. Software can be sold to millions with very little effort by the smallest of businesses, unlike physical products that need an enormous amount of investment for things like factories and such. Unless you're stupid enough to make your invention public before you sell it, you'll be firmly planted within your market within days. It'll be over a year or so before a big business has it in their product.

        Large businesses tend to be slow and plodding in the software market. Small businesses and individuals tend to be agile and fast. This is opposite of the conditions needed to make your claim true.
      • RE: Spend your money/time developing some software product...

        Like Angry birds? Facebook? Google? Apple? Microsoft? They were all small businesses when they came out. How has software patents protected them?

        A good example of being protected by patents is I4I. Where are they now?
    • How about a fourth position: make the patent office follow its own rules

      I don't have a problem with software patents, if they were awarded for things that were truly inventive and non-obvious. If you did that there would be vastly fewer patents awarded, and they would represent, protect, and encourage true innovation.

      If you look at a lot of the most egregious software patents, they are for things that completely violate the 'non-obvious' requirement. For example, they are issued for things that people do all the time with physical objects, but somehow applying them to a virtual or software defined object suddenly makes that a 'new' invention. I don't think so. For instance, could you patent a lid for a box with a sliding clasp that you push with your finger to open? Heavens no, latches like that have existed for thousands of years. Yet Apple was somehow awarded a patent for a sliding latch to open your cell phone screen. That's inscrutable. Or how about Amazon's 'one-click' buy button. How on earth is that a non-obvious or innovative idea?

      If you simply threw out all of the patents that have been issued that, under the longstanding actual rules of the patent office should never have been issued, then we probably wouldn't have the uproar and controversy and ridiculous situations we have. Simply require the patent office to stick to its own actual rules.
      • RE: Simply require the patent office to stick to its own actual rules

        It's not that simple. The USPTO is a self-funded entity. However, the U.S. Congress with it's insatiable need to stuff money down various rat holes has raided the USPTO cookie jar:


        Thus, making it more difficult for the USPTO to hang onto experienced staff as well as hire new employees to meet increasing demand.

        The America Invents Act, passed by Congress and signed by the President in 2011, did not include language that would bar Congress from appropriating USPTO funds.
        Rabid Howler Monkey
  • Software is much more like a machine?

    I've written software for 30 some years. From the patent perspective, your statement is fundamentally wrong. To build, manufacture, and distribute a machine takes deep resources and quite significant [b]investment[/b]. Patents in this case are protecting people who are making said investment so they actually feel it's worth pursuing. I've said it before in comments to your assertions, if I can sit at my home office desk and using my own ideas, my own code, and a simple computer, re-create said "invention", [b]and[/b] distribute it over the internet, albeit not as efficiently as a big company, there is no way you can argue it's like a machine, nor that it deserves the same protection.

    And, as a small independent developer, even if I'm first out the door, I can't afford the lawyers and money needed to gain such a patent, leaving me open to being taken to task by some big company who can afford it. Furthermore, I can't afford to even defend myself in that case, even if it's just to prove prior art (especially since the patent office seems incapable or unwilling to find prior art). So, patents work [i]against[/i] the so-called rightful owner in that case.
    • I got yer research right here

      Your second scenario is easily dealt with. Just start a blog called "oldCoderException's Journal of Research" where you publish all the stuff you thought up but can't afford to patent. IBM (and Bell Labs, back when they existed) published "Journals of Research" in which they put all the stuff they didn't want to bother patenting, but which they didn't want anyone else patenting either.
      Robert Hahn
      • You missed my point re: defending myself

        My point was not that I did not have prior art (heck, if I distributed it on the internet, I have that). My point was, and remains, that I'd be hard pressed or find it impossible to defend myself against a massive company who thinks they own a patent on my "so called" invention. And even if you have said prior art, proving it can be harder than you may think on the surface.
    • Are you for or against?

      Patents are NOT concerned with how easily something can be manufactured or distributed. Trying to say that software isn't like a machine because it is easier to do those things is missing the point.

      A patent is about being able to create an independent manifestation of the idea. The exact medium of that manifestation, means of producing it or means of distributing it are not relevant.
  • I still don't understand how

    it is legal to patent an arbitrary shape. Let alone how and why an agency could even allow such foolishness, though it seems a modern thing. Similar patent attempts in the past have always failed in court, just not now. Sorry for the variance from the topic, but I'm just a greasemonkey, not a coder.
  • All USA-only software patents do is make everything more expensive

    They need to be abolished. Everyone is infringing by default and that's not something you should be worried about if you copy someone else's idea, but write your own code from scratch. Damages are too high and the life of the patent is too long. Software patents are a mess in the USA! Lawyers, patent trolls and companies who want to temporarily slow down competition use patents as anti-competitive business weapons. Software patents only benefit a few at a great expense to others and the rest of the developed world seems to be doing fine without them as was the USA before they were fully legitimized in 1998. Microsoft built it's fortune in a software-patent free world...
    • Destroy or Hoard

      Software patents today are primarily being used as swords and shields not as hammer and anvils.
  • Software and ideas

    Where is the boundary between software and ideas? Would a textbook on algorithms be found to infringe on patents if they describe a patented algorithm or process? Any software can be mathematically described. I can show that a pay at the pump gas station infringes on Amazon's One click patent.

    Examine the patents being asserted, an interface that is handed a message to sends that automatically splits it up into smaller pieces because of a lower layer limitation is violating a patent. A few lines of code to do make some action occur when an image is moved on top of another image is violating a patent. A few lines of code to put a context menu on a hyperlink that is different then the surrounding text is violating a patent. A few lines of code to apply a regular expression to a text to identify email addresses and insert a mail to tag violates another patent. A spell checker that identifies misspelled words and offers the user a choice of actions is violating a patent.

    With the advances in programming, how can I not violate some patent every day? Maybe even every 5 minutes? How can I understand, let alone remember millions of patents?

    That is why I am against software patents. It is very easy for me to transfer ideas to software. Does that mean my brain is in violation of all these patents? At least it enables me to find prior art easily.
    • Deleted ...........posted wrong level..............