How Not to Argue for Software Patents

by on January 17, 2006 · 22 comments

Is Patrick Ross trying to make his own side look bad?

The conference heard a compelling perspective on software patents from Guenther Schmalz, director of IP for Europe for SAP, the largest European software maker and the third-largest software maker in the world. SAP burst on the scene in the early1970s, like Microsoft moving into areas of software that IBM had no interest in commoditizing. For years SAP grew and grew and had no patents. Now they’re ramping up a major patent office with a few dozen attorneys around the world. Why? “Times have changed,” Schmalz said. In the early days, SAP’s competitors were way behind, but now competitors are everywhere. Patents are the only way for SAP to ensure returns on its development investment, he said, adding that copyright is no solution, as the actual writing of code only makes up about 20% of the development of software.

“Those who drive innovation need patents,” he said. “Those who don’t imitate.”

To put this a little bit differently, in the past, SAP was an innovative company that was able to stay ahead of the competition by virtue of their superior technology. However, now that they’re a fat, lazy incumbent, they’re discovering the joys of using patent law as a club against their more innovative competitors.

The idea that creators have the right to recoup all of the value they create is a common idea in intellectual property debates. Larry Lessig quotes NYU law professor Rochelle Dreyfuss as dubbing this the “if value, then right” theory of intellectual property. It’s an idea that sounds intuitively plausible, but with a little bit of reflection, it becomes obvious how pernicious it is.

After all, capitalism is founded on the ideal of vigorous competition. When some entrepreneur invents a new product category of business model–say, the fast-food restaurant–he invariably attracts a swarm of competitors. Competition forces down the prices in this newly-created industry, preventing the innovator from capturing the full value of his or her invention. Ray Kroc’s invention of the fast food restaurant, for example, created a whole lot of value, and most of that value went to consumers, not to Kroc. He wasn’t able to use the law to exclude competitors and “ensure returns” on his innovation. To the contrary, under our economic system, he was required to continue innovating if he wanted to turn a profit.

It’s quite possible that Kroc felt resentful toward these upstart “imitators” who copied his business model without (he would claim) adding value themselves. But that’s the way the business world works. Competition is cutthroat, and businesses who stop innovating shrivel up and die. That’s as it should be, and consumers benefit tremendously from such creative destruction.

By the same token, SAP probably resents all these pesky little companies that are barging in on market opportunities that they pioneered. Tough. If they want to stay at the top of their industry, they need to continue innovating. They have no right to expect the government to “ensure returns” on their investments. That’s the way it goes in most other industries, and it’s the way the software industry ought to work as well.

  • Nabil

    I agree with you in principle. Most software patents are actually a bit absurd . The patent system for software is beyond broken. Probably for other industries too but I used to work in software so I can speak to that.

    However, most big software companies have traditionally used patents defensively to protect against getting sued instead of as a weapon to squash competitors. Because everyone holds so many patents on such obscure little things it is almost impossible not to be violating someone elses patents. So if Oracle sued SAP for violating a patent, SAP would just find a few that Oracle was violating and counter sue. They would settle, pay each other off and the lawyers get to make lots of money (or work absurdly long hours if you are an in house lawyer). A far bigger impediment to competition in software markets like SAPs is a lack of standards and extremely complex software that the installed base cant rip out.

    I also dont really think it is acurate to say “thats the way it goes in most other industries”. It seems that most other intellectual property industries have extensive patent and copyright protections.

    One last question: Is the graphic at the top intended to invoke thoughts of the soviet union? Or is that just me?

  • Nabil

    I agree with you in principle. Most software patents are actually a bit absurd . The patent system for software is beyond broken. Probably for other industries too but I used to work in software so I can speak to that.

    However, most big software companies have traditionally used patents defensively to protect against getting sued instead of as a weapon to squash competitors. Because everyone holds so many patents on such obscure little things it is almost impossible not to be violating someone elses patents. So if Oracle sued SAP for violating a patent, SAP would just find a few that Oracle was violating and counter sue. They would settle, pay each other off and the lawyers get to make lots of money (or work absurdly long hours if you are an in house lawyer). A far bigger impediment to competition in software markets like SAPs is a lack of standards and extremely complex software that the installed base cant rip out.

    I also dont really think it is acurate to say “thats the way it goes in most other industries”. It seems that most other intellectual property industries have extensive patent and copyright protections.

    One last question: Is the graphic at the top intended to invoke thoughts of the soviet union? Or is that just me?

  • http://www.binarybits.org/ Tim

    Nabil,

    I certainly don’t blame big software companies for their defensive patenting practices. I would probably do likewise in their shoes. But what I’m interested in is the policy question of whether patents are good for innovation in the software industry. The contention that Patrick Ross seems to be making–that SAP needs software patents to prevent competitors from imitating them–seems to me to profoundly misunderstand the nature of competition and innovation in a free market.

    And when I say “that’s the way it goes in most other industries,” I didn’t mean to say that there are no patents in those industries. My point was that in most industries, companies don’t expect the law to give them ways to prevent their competitors from imitating them. Yet that appears to be precisely what SAP is arguing for in the quote above.

  • http://www.binarybits.org/ Tim

    Nabil,

    I certainly don’t blame big software companies for their defensive patenting practices. I would probably do likewise in their shoes. But what I’m interested in is the policy question of whether patents are good for innovation in the software industry. The contention that Patrick Ross seems to be making–that SAP needs software patents to prevent competitors from imitating them–seems to me to profoundly misunderstand the nature of competition and innovation in a free market.

    And when I say “that’s the way it goes in most other industries,” I didn’t mean to say that there are no patents in those industries. My point was that in most industries, companies don’t expect the law to give them ways to prevent their competitors from imitating them. Yet that appears to be precisely what SAP is arguing for in the quote above.

  • Geoff

    There’s as big a flaw in your reasoning that competition is good as there is that software patents stifle competition.

    As you point out:
    “Competition is cutthroat, and businesses who stop innovating shrivel up and die.”

    What about a small start-up software company who make their great new piece of software and try to sell it, only to have some big player (Microsoft, say) copy the ideas and then integrate them for free in their existing products? This is great for the consumer (so long as the consumer wants nothing more than cheap products) but the bigger, badder, corporation can quite easily kill the smaller competition. Microsoft have been doing this for years and despite numerous attempts to have them done for anti-competitive behaviour, they’ve hardly been scratched by the law suits. A completely free-market is never going to be the solution.

    In truth, patents ARE anti-competitive. This is widely recognised. Nevertheless, patents have existed in all fields of technology for several centuries and have enabled the ideas people to get a head start in using their ideas before everyone copies them. This is their reward, and I see no reason why such people shouldn’t be rewarded. Yet patents can also be damaging, so a balance must always be found between their inherently anti-competitive nature and the reward that is given to the inventor.

    Looking at the practice in the US where they have:
    1. Piss poor searching for (and application of) prior art;
    2. Assumption of validity of a granted patent, combined with virtually useless re-examination system mean that you have to go to court to fight your case, and you’re always fighting an uphill battle;
    3. Punitive triple damages, making patents a weapon rather than a shield

    I think that it is easy to see that the balance is all wrong. Patents, in any field of technology, are too powerful in the US and this is what’s damaging, not that software patents exist.

    In Europe, things are quite different and while I can’t say whether the balance is right, it’s certainly better:
    1. No assumption of validity of a granted patent.
    2. Slightly better searching of the prior art combined with the requirement for a “technical effect” meaning that the Europeans don’t even have to try looking for prior art to show that it’s obvious to use a computer to remind you when to walk your dog.
    3. Opportunities for opposing granted patents at the Patent Office directly. This helps correct for mistakes, which every patent office will make – even the EPO granted a patent based on Amazon’s “one-click” application (although not for that particular invention) which is currently under serious opposition from several parties.
    4. Non-punitive damages.

    Nevertheless, court costs are still high, which protects the big business over the small inventor. Maybe more should be done in Europe and the US to combat this problem.

    I would also point out that the software field is a lot different today than it was in the 1970′s not just because of software patents, but also because copying software is immeasurably easier. You could argue that software patents are a necessary change to combat the ever-increasing threat of piracy.

    What I’m trying to say is that software patents are only evil to the extent that all patents are evil. But their perniciousness lies not in the fact that they exist, but in that the power afforded to patent-wielders is too high.

  • Geoff

    There’s as big a flaw in your reasoning that competition is good as there is that software patents stifle competition.

    As you point out:
    “Competition is cutthroat, and businesses who stop innovating shrivel up and die.”

    What about a small start-up software company who make their great new piece of software and try to sell it, only to have some big player (Microsoft, say) copy the ideas and then integrate them for free in their existing products? This is great for the consumer (so long as the consumer wants nothing more than cheap products) but the bigger, badder, corporation can quite easily kill the smaller competition. Microsoft have been doing this for years and despite numerous attempts to have them done for anti-competitive behaviour, they’ve hardly been scratched by the law suits. A completely free-market is never going to be the solution.

    In truth, patents ARE anti-competitive. This is widely recognised. Nevertheless, patents have existed in all fields of technology for several centuries and have enabled the ideas people to get a head start in using their ideas before everyone copies them. This is their reward, and I see no reason why such people shouldn’t be rewarded. Yet patents can also be damaging, so a balance must always be found between their inherently anti-competitive nature and the reward that is given to the inventor.

    Looking at the practice in the US where they have:
    1. Piss poor searching for (and application of) prior art;
    2. Assumption of validity of a granted patent, combined with virtually useless re-examination system mean that you have to go to court to fight your case, and you’re always fighting an uphill battle;
    3. Punitive triple damages, making patents a weapon rather than a shield

    I think that it is easy to see that the balance is all wrong. Patents, in any field of technology, are too powerful in the US and this is what’s damaging, not that software patents exist.

    In Europe, things are quite different and while I can’t say whether the balance is right, it’s certainly better:
    1. No assumption of validity of a granted patent.
    2. Slightly better searching of the prior art combined with the requirement for a “technical effect” meaning that the Europeans don’t even have to try looking for prior art to show that it’s obvious to use a computer to remind you when to walk your dog.
    3. Opportunities for opposing granted patents at the Patent Office directly. This helps correct for mistakes, which every patent office will make – even the EPO granted a patent based on Amazon’s “one-click” application (although not for that particular invention) which is currently under serious opposition from several parties.
    4. Non-punitive damages.

    Nevertheless, court costs are still high, which protects the big business over the small inventor. Maybe more should be done in Europe and the US to combat this problem.

    I would also point out that the software field is a lot different today than it was in the 1970′s not just because of software patents, but also because copying software is immeasurably easier. You could argue that software patents are a necessary change to combat the ever-increasing threat of piracy.

    What I’m trying to say is that software patents are only evil to the extent that all patents are evil. But their perniciousness lies not in the fact that they exist, but in that the power afforded to patent-wielders is too high.

  • http://www.angryblog.org/ Tim

    I don’t see anything wrong with the way the software industry has been working vis-a-vis Microsoft and small startups. When you decide to launch a new software product, one of the risks of doing business is that Microsoft might come along and build a competing product. But that doesn’t always happen. And even when it does happen, they don’t always succeed.

    Indeed, what often happens is that instead of building a competing product from scratch, Microsoft will buy out the existing business and integrate its software into its products. See, for example, Bungie, Connectix, and Hotmail. Apple does the same thing. This is good for innovation because it increases the incentive to create start-ups: getting bought out by Microsoft can be quite lucrative.

    I agree that there’s probably a need to reform patent law in general in the United States. However, I also think that software patents are particularly pernicious. The reason is that I think that virtually all software “inventions” are obvious, because computer programming just isn’t a capital-intensive activity. I make this argument in more details here.

  • http://www.angryblog.org/ Tim

    I don’t see anything wrong with the way the software industry has been working vis-a-vis Microsoft and small startups. When you decide to launch a new software product, one of the risks of doing business is that Microsoft might come along and build a competing product. But that doesn’t always happen. And even when it does happen, they don’t always succeed.

    Indeed, what often happens is that instead of building a competing product from scratch, Microsoft will buy out the existing business and integrate its software into its products. See, for example, Bungie, Connectix, and Hotmail. Apple does the same thing. This is good for innovation because it increases the incentive to create start-ups: getting bought out by Microsoft can be quite lucrative.

    I agree that there’s probably a need to reform patent law in general in the United States. However, I also think that software patents are particularly pernicious. The reason is that I think that virtually all software “inventions” are obvious, because computer programming just isn’t a capital-intensive activity. I make this argument in more details here.

  • patent troll

    “I think that virtually all software “inventions” are obvious, because computer programming just isn’t a capital-intensive activity” said Tim

    This is just one completely illogical ( not to say downright foolish) statement of yours.
    I guess you are one of those Slashbot-type folks who honestly think that software is just written, not engineered.
    Well, go ahead and write a lot of your junky software. You sure don’t need patents to protect it.
    Patents are needed to protect gems like RSA algorithm, not the junk people like you come up with. Copyright is all you need…

  • patent troll

    “I think that virtually all software “inventions” are obvious, because computer programming just isn’t a capital-intensive activity” said Tim

    This is just one completely illogical ( not to say downright foolish) statement of yours.
    I guess you are one of those Slashbot-type folks who honestly think that software is just written, not engineered.
    Well, go ahead and write a lot of your junky software. You sure don’t need patents to protect it.
    Patents are needed to protect gems like RSA algorithm, not the junk people like you come up with. Copyright is all you need…

  • Jack

    Nabil,

    Defensive patenting only works for big companies who have extensive patent portfolios. And even there, we are beginning to see how harmful the practice is. On the side of competition and free-market-based pricing, we see Microsoft beginning to demand royalties for some of their patents (a FAT filesystem tax for all makers of USB storage devices, digital cameras, etc., not to mention the possibility of MS stomping on Linux’s ability to read these devices because of their FAT patent). On the side of the big companies, defensive patents don’t do a thing against patent trolls, who make no product whatsoever and thus can’t be counter-sued using a defensive patent portfolio.

    Tim’s arguments about free-markets should also be a key to understanding that he isn’t invoking soviet russia. Quite the opposite, in fact — patents in general more closely resemble Communist Central Planning and banning of “free-market inefficiencies” than does the banner graphic (though I agree that it invoke some similarities graphically).

  • Jack

    Nabil,

    Defensive patenting only works for big companies who have extensive patent portfolios. And even there, we are beginning to see how harmful the practice is. On the side of competition and free-market-based pricing, we see Microsoft beginning to demand royalties for some of their patents (a FAT filesystem tax for all makers of USB storage devices, digital cameras, etc., not to mention the possibility of MS stomping on Linux’s ability to read these devices because of their FAT patent). On the side of the big companies, defensive patents don’t do a thing against patent trolls, who make no product whatsoever and thus can’t be counter-sued using a defensive patent portfolio.

    Tim’s arguments about free-markets should also be a key to understanding that he isn’t invoking soviet russia. Quite the opposite, in fact — patents in general more closely resemble Communist Central Planning and banning of “free-market inefficiencies” than does the banner graphic (though I agree that it invoke some similarities graphically).

  • Jack

    patent troll,

    It is obvious that you either a) don’t know a lot about software engineering or b) have a self-inflated idea about how hard it is.

    The truth is, writing good software is not harder than writing a good novel, in terms of investment in time and resources. Sure, absolutely — software takes some specialized knowledge that isn’t generally rammed into students heads K-12 like writing is, but that doesn’t change the essential nature of the thing: any motivated person can sit down with a pencil and paper and write a novel, and any motivated person can sit down with a keyboard and mouse and write a computer program.

    I suppose you would support granting patents on novels, plotlines, and other literary devices. But most of us find that absurd. And if you were familiar with software, you’d see that the absurdity is analagous — patenting simple algorithms, data structures, user-interface elements, combinations of existing software, etc. are no different than patenting a few characteristics of a character in a novel, the interactions between characters and objects, the flow of time, twists in the plot, novel use of language, etc.

    And, lest you get the wrong idea, I’m a full time software developer and researcher. I love writing software and creating new things. I do it not only at work, but in my spare time. I’ve written published articles on it. I just don’t happen to think that my passion and knowledge should make me any more special than any other creator, and that we shouldn’t be giving 20-year monopolies to people for obvious and simple ideas.

  • Jack

    patent troll,

    It is obvious that you either a) don’t know a lot about software engineering or b) have a self-inflated idea about how hard it is.

    The truth is, writing good software is not harder than writing a good novel, in terms of investment in time and resources. Sure, absolutely — software takes some specialized knowledge that isn’t generally rammed into students heads K-12 like writing is, but that doesn’t change the essential nature of the thing: any motivated person can sit down with a pencil and paper and write a novel, and any motivated person can sit down with a keyboard and mouse and write a computer program.

    I suppose you would support granting patents on novels, plotlines, and other literary devices. But most of us find that absurd. And if you were familiar with software, you’d see that the absurdity is analagous — patenting simple algorithms, data structures, user-interface elements, combinations of existing software, etc. are no different than patenting a few characteristics of a character in a novel, the interactions between characters and objects, the flow of time, twists in the plot, novel use of language, etc.

    And, lest you get the wrong idea, I’m a full time software developer and researcher. I love writing software and creating new things. I do it not only at work, but in my spare time. I’ve written published articles on it. I just don’t happen to think that my passion and knowledge should make me any more special than any other creator, and that we shouldn’t be giving 20-year monopolies to people for obvious and simple ideas.

  • Nabil

    Jack,

    Point taken about patent trolls, there should be a special place in hell for them. And, as I said in my post, I do fundamentally agree that the software patent system is broken. The way the patents are granted around such tiny minutia is ridiculous. But I do think there is a place for patents in software. If for nothing else than to help small software companies protect against the big guys stealing IP after in-depth technical demos or partnership/acquisition related due diligence. I just believe the bar needs to be MUCH higher to get a patent approved. As I re-read Tim’s post (and since I am new to the site) I am not sure if he is against all patents in software or just the state of the patent system as it is today.

    As for Microsoft, well they are Microsoft, what can I say. And while I have been out of software for a few months I still think the vast majority of software companies don’t use patents aggressively but no doubt with the way the system is today that could change.

    On the graphic, the comment was partially in jest. But it is very visually similar to soviet images, including that famous propaganda play about the rights of the worker (that I can’t for the life of me remember the name of). I did not get the connection between the graphic and the content on the site (especially since the Tim used to be with the Cato Institute) so I was just wondering if it was a coincidence or there was some point.

  • Nabil

    Jack,

    Point taken about patent trolls, there should be a special place in hell for them. And, as I said in my post, I do fundamentally agree that the software patent system is broken. The way the patents are granted around such tiny minutia is ridiculous. But I do think there is a place for patents in software. If for nothing else than to help small software companies protect against the big guys stealing IP after in-depth technical demos or partnership/acquisition related due diligence. I just believe the bar needs to be MUCH higher to get a patent approved. As I re-read Tim’s post (and since I am new to the site) I am not sure if he is against all patents in software or just the state of the patent system as it is today.

    As for Microsoft, well they are Microsoft, what can I say. And while I have been out of software for a few months I still think the vast majority of software companies don’t use patents aggressively but no doubt with the way the system is today that could change.

    On the graphic, the comment was partially in jest. But it is very visually similar to soviet images, including that famous propaganda play about the rights of the worker (that I can’t for the life of me remember the name of). I did not get the connection between the graphic and the content on the site (especially since the Tim used to be with the Cato Institute) so I was just wondering if it was a coincidence or there was some point.

  • http://www.angryblog.org/ Tim

    Nabil,

    I think the Soviet-style imagery is intended to be kitchy and ironic. 20th century “liberation fronts” have traditionally been Marxist in orientation, but we’re a “liberation front” for free markets and individual liberty.

    Oh, and to answer your question, I do tend to think that the world would be a better place without software patents, although I do admit there are hard cases like RSA where the question is debatable.

  • http://www.angryblog.org/ Tim

    Nabil,

    I think the Soviet-style imagery is intended to be kitchy and ironic. 20th century “liberation fronts” have traditionally been Marxist in orientation, but we’re a “liberation front” for free markets and individual liberty.

    Oh, and to answer your question, I do tend to think that the world would be a better place without software patents, although I do admit there are hard cases like RSA where the question is debatable.

  • Ed

    As a software engineer and aspiring author, I think I’m fairly well qualified to state which is easier, software development, or novel writing.

    Software development is easier, hands down, no contest. When one is developing software, and following good programming practices (making frequent use of revision control, unit tests, functional tests, and profiling), it’s difficult to make a really huge mistake. There may be some question about when you’ll finish, but one can make incremental improvements. Any error in the process can easily be backed out to a known good state. If the error is missed for several itterations, a good revision control program can potentially ease the burden to fix the error without losing any good code produced since then – so long as, of course, it doesn’t depend on the error.

    When one is writing a novel, one has none of these aids. True, one can use revision control – as a software developer/author, I use revision control in my writing. However, it doesn’t really help much – the biggest benefit of revision control in novel writing is figuring out what was mis-edited when one decided to go back and add some foreshadowing. There’s no debuggers, no profilers for novels. The nearest equivalent are friends and family who are willing to read ones unrefined drivel. Depending on their number, dedication, and capabilities, an aspiring author may be severely restricted in the number of proofreads available; in any event, they are very time-consuming, subjective, and otherwise of limited use. Even with infinite proofreading resources, there’s no guarantee that the novel will be up to performing its intended function.

    Collaboration… adding several programmers to a project, with sufficient coordination, will generally hasten its development. If the project is large enough, and can be broken into sufficient sub-parts, hundreds, or even thousands of programmers can work together on it, without appreciable inefficiency (assuming, as always, communication and coordination.) For example, the average Linux distribution includes the work output of tens of thousands of programmers – and most of that is actually without coordination. (The Linux kernel is coordinated. But there isn’t much coordination between the kernel developers and bash or xterm or other non-core components.)

    Novels written by teams of writers tend to work in one of two fashions: one writer dominates, acting as an editor for the work as a whole; poorly. It doesn’t scale, so long as there is to be any flow between the pieces.

  • Ed

    As a software engineer and aspiring author, I think I’m fairly well qualified to state which is easier, software development, or novel writing.

    Software development is easier, hands down, no contest. When one is developing software, and following good programming practices (making frequent use of revision control, unit tests, functional tests, and profiling), it’s difficult to make a really huge mistake. There may be some question about when you’ll finish, but one can make incremental improvements. Any error in the process can easily be backed out to a known good state. If the error is missed for several itterations, a good revision control program can potentially ease the burden to fix the error without losing any good code produced since then – so long as, of course, it doesn’t depend on the error.

    When one is writing a novel, one has none of these aids. True, one can use revision control – as a software developer/author, I use revision control in my writing. However, it doesn’t really help much – the biggest benefit of revision control in novel writing is figuring out what was mis-edited when one decided to go back and add some foreshadowing. There’s no debuggers, no profilers for novels. The nearest equivalent are friends and family who are willing to read ones unrefined drivel. Depending on their number, dedication, and capabilities, an aspiring author may be severely restricted in the number of proofreads available; in any event, they are very time-consuming, subjective, and otherwise of limited use. Even with infinite proofreading resources, there’s no guarantee that the novel will be up to performing its intended function.

    Collaboration… adding several programmers to a project, with sufficient coordination, will generally hasten its development. If the project is large enough, and can be broken into sufficient sub-parts, hundreds, or even thousands of programmers can work together on it, without appreciable inefficiency (assuming, as always, communication and coordination.) For example, the average Linux distribution includes the work output of tens of thousands of programmers – and most of that is actually without coordination. (The Linux kernel is coordinated. But there isn’t much coordination between the kernel developers and bash or xterm or other non-core components.)

    Novels written by teams of writers tend to work in one of two fashions: one writer dominates, acting as an editor for the work as a whole; poorly. It doesn’t scale, so long as there is to be any flow between the pieces.

  • http://www.abc-acupuncture.com/baxqorav tramadol

    81e31de21f46 Hello! http://www.abc-acupuncture.com/baxqorav tramadol

  • http://www.abc-acupuncture.com/baxqorav tramadol

    81e31de21f46 Hello! http://www.abc-acupuncture.com/baxqorav tramadol

Previous post:

Next post: