MPEG Patent Thicket

by on May 11, 2006 · 24 comments

In comments to my big DVD post, Urijah points out another obstacle to a completely free and legal version of MPlayer or Xine: the MPEG format is heavily encumbered by patents, and commercial entities generall must pay $2.50 per installation for a license.

I haven’t looked into this issue in detail, but if this article is right, this problem likely extends beyond MPEG-2 to other video-playback technologies:

All patents in the list of the MPEG licence association in regard to the MPEG-4 standard were examined and analysed. After intensive study of relevant literature and more than 100 patents of the relevant companies we can say now: Upon careful examination, we can not find any advances over the prior art in said list that could justify the granting of a patent. Most of these patents should be attackable in court, but who would take the burden of litigation against 900 patents owned by dozens of large companies?

When people talk about “the MPEG-2 patents,” they aren’t referring to a specific patent that describes the MPEG-2 standard. Rather, they refer to 640 holders of patents related to various aspects of the MPEG-2 standard. If 640 patents describe a single video format, it’s a safe bet that a substantial fraction of them cover any conceivable alternative video format. Which means that technically speaking, all free video-playing software is probably technically infringing on numerous patents.

It’s also quite possible that the MPEG-LA wouldn’t bother suing an open source project, which doesn’t have any money anyway. At worst, I would think MPlayer and Xine could charge people $2.50 to download copies of MPlayer or Xine and turn the tribute revenue over to the MPEG-LA. They could still distribute the source code, so this wouldn’t greatly hamper its development as an open source project.

In any event, the best solution is to repeal software patents, which impede innovation in this and many other software categories.

  • Lewis Baumstark

    I think we need to be careful about repeating the “repeal software patents” mantra. While I can’t think of a single software patent that is (in my mind) legit, I’m hesitant to throw them out entirely. (Mostly because it seems a band-aid solution.) As a software developer (well, as someone who *teaches* software development) I’m at least willing to concede that a few things in that domain are patentable. It seems to me that if the bar for patentability was set appropriately high, most software patents would never be approved. And rightly so.

  • http://www.blindmindseye.com MikeT

    The USPTO should not be funded by user fees. That is just asking for corruption and incompetence. They need to be able to hire bonafide practitioners with strong educations in these subjects with very good salaries. Until that changes, I don’t think the bar can be set high because the average examiner won’t even understand the bar.

  • http://www.techliberation.com/ Tim

    Lewis,

    Can you give me some examples of legitimate software patents? I’ve seen dozens of examples of bogus software patents, but I’ve yet to find a really compelling example of a software patent that promoted innovation.

    I also have very little confidence in the ability of the patent office to distinguish between legitimate and illegitimate software patents. If (say) 98% of software patents are bogus and 2% are legitimate, even a relatively low error rate will cause the overwhelming majority of software patents to be bogus. It’s fine to say that the bar should be raised, but I’m not sure it’s possible to do so unless the patent office hires some people with much greater technical skills.

  • Lewis Baumstark

    I think we need to be careful about repeating the “repeal software patents” mantra. While I can’t think of a single software patent that is (in my mind) legit, I’m hesitant to throw them out entirely. (Mostly because it seems a band-aid solution.) As a software developer (well, as someone who *teaches* software development) I’m at least willing to concede that a few things in that domain are patentable. It seems to me that if the bar for patentability was set appropriately high, most software patents would never be approved. And rightly so.

  • http://www.blindmindseye.com MikeT

    The USPTO should not be funded by user fees. That is just asking for corruption and incompetence. They need to be able to hire bonafide practitioners with strong educations in these subjects with very good salaries. Until that changes, I don’t think the bar can be set high because the average examiner won’t even understand the bar.

  • http://www.techliberation.com/ Tim

    Lewis,

    Can you give me some examples of legitimate software patents? I’ve seen dozens of examples of bogus software patents, but I’ve yet to find a really compelling example of a software patent that promoted innovation.

    I also have very little confidence in the ability of the patent office to distinguish between legitimate and illegitimate software patents. If (say) 98% of software patents are bogus and 2% are legitimate, even a relatively low error rate will cause the overwhelming majority of software patents to be bogus. It’s fine to say that the bar should be raised, but I’m not sure it’s possible to do so unless the patent office hires some people with much greater technical skills.

  • http://www.ssokolow.com/ Stephan Sokolow

    Of course, if they had to pay royalties, then the open-source world would suffer because any license which made illegal any concerted effort to circumvent the license fee (eg. mirroring, cacheing, or even casual sharing) would be in conflict with the source availability terms in the GNU GPL license and therefore the MPlayer source code couldn’t be linked against any GPLed libraries.

    I haven’t checked, but it’s very likely that MPlayer does gain at least two major features from 3rd-party GPLed code.

  • Urijah

    At worst, I would think MPlayer and Xine could charge people $2.50 to download copies of MPlayer or Xine and turn the tribute revenue over to the MPEG-LA.

    I don’t see how that would square with the GPL–MPEG-LA would not allow these programs to have a license that said you were free to make copies for your friends, otherwise only one person would pay for the program and then distribute it. Oh, and you have to pay AC-3 royalties too.

  • Lewis Baumstark

    Can you give me some examples of legitimate software patents? … I’ve yet to find a really compelling example of a software patent that promoted innovation.

    Neither have I, this was why I said I couldn’t come up with anything. My point is that I’m at least willing to believe there are a few out there, despite all the bad ones. My fear is that tossing out *all* software patents would have unforseen consequences ala net neutrality legislation, DMCA, etc.

    Regarding a “good” software patent: my first thought was the Fast Fourier Transform (FFT), but it turns out it wasn’t patented (I thought it was, but in fact, if the rumors are true, the authors published it to keep it from being patented). Still, it seems to me to come close to what is, in my head, a patentable software entity, mainly that it is a specific implementation (typically patentable) of a general concept (ideally non-patentable). And it certainly was something that promoted innovation (hand me a multimedia device and odds are it uses an FFT somewhere); it was so important that I doubt patenting it would have held up that innovation.

    It’s fine to say that the bar should be raised, but I’m not sure it’s possible to do so unless the patent office hires some people with much greater technical skills.

    Preaching to the choir here, except that it seems to me this reaches beyond software into other areas (business method patents, for example).

  • http://www.ssokolow.com/ Stephan Sokolow

    Of course, if they had to pay royalties, then the open-source world would suffer because any license which made illegal any concerted effort to circumvent the license fee (eg. mirroring, cacheing, or even casual sharing) would be in conflict with the source availability terms in the GNU GPL license and therefore the MPlayer source code couldn’t be linked against any GPLed libraries.

    I haven’t checked, but it’s very likely that MPlayer does gain at least two major features from 3rd-party GPLed code.

  • Urijah

    At worst, I would think MPlayer and Xine could charge people $2.50 to download copies of MPlayer or Xine and turn the tribute revenue over to the MPEG-LA.

    I don’t see how that would square with the GPL–MPEG-LA would not allow these programs to have a license that said you were free to make copies for your friends, otherwise only one person would pay for the program and then distribute it. Oh, and you have to pay AC-3 royalties too.

  • http://www.techliberation.com/ Tim

    Stephan,

    Don’t a lot of GPLed libraries use the LGPL to allow proprietary software to link against them?

  • http://www.techliberation.com/ Tim

    Lewis,

    In case you haven’t seen it, I give a couple of reasons here that I think software patents are uniquely bad. One, software, unlike most other inventions, already enjoys copyright protection. Two, the capital investment per “invention” is much lower for software than for other categories of patents, such as pharmaceuticals. Hence, software patents are much more likely to be trivial. And three, because programming typically involves applying well-known software engineering techniques to new problems, virtually all software inventions are “obvious” in the sense that any competent programmer would be able to solve a given problem.

    I’m hoping to do more in-depth research on the subject at some point, and I might find examples that change my mind, but at the moment, disallowing software patents entirely seems like the cleanest way to reform the current mess.

  • Lewis Baumstark

    Can you give me some examples of legitimate software patents? … I’ve yet to find a really compelling example of a software patent that promoted innovation.

    Neither have I, this was why I said I couldn’t come up with anything. My point is that I’m at least willing to believe there are a few out there, despite all the bad ones. My fear is that tossing out *all* software patents would have unforseen consequences ala net neutrality legislation, DMCA, etc.

    Regarding a “good” software patent: my first thought was the Fast Fourier Transform (FFT), but it turns out it wasn’t patented (I thought it was, but in fact, if the rumors are true, the authors published it to keep it from being patented). Still, it seems to me to come close to what is, in my head, a patentable software entity, mainly that it is a specific implementation (typically patentable) of a general concept (ideally non-patentable). And it certainly was something that promoted innovation (hand me a multimedia device and odds are it uses an FFT somewhere); it was so important that I doubt patenting it would have held up that innovation.

    It’s fine to say that the bar should be raised, but I’m not sure it’s possible to do so unless the patent office hires some people with much greater technical skills.

    Preaching to the choir here, except that it seems to me this reaches beyond software into other areas (business method patents, for example).

  • http://www.techliberation.com/ Tim

    Stephan,

    Don’t a lot of GPLed libraries use the LGPL to allow proprietary software to link against them?

  • http://www.techliberation.com/ Tim

    Lewis,

    In case you haven’t seen it, I give a couple of reasons here that I think software patents are uniquely bad. One, software, unlike most other inventions, already enjoys copyright protection. Two, the capital investment per “invention” is much lower for software than for other categories of patents, such as pharmaceuticals. Hence, software patents are much more likely to be trivial. And three, because programming typically involves applying well-known software engineering techniques to new problems, virtually all software inventions are “obvious” in the sense that any competent programmer would be able to solve a given problem.

    I’m hoping to do more in-depth research on the subject at some point, and I might find examples that change my mind, but at the moment, disallowing software patents entirely seems like the cleanest way to reform the current mess.

  • http://todd.tv Todd

    There are a couple of other really good reasons why software patents are terrible policy:

    1. The ‘borders’ of a software patent are unclear.

    2. Software patents are detrimental to startups.

    3. Software patents in the U.S. are causing us to ship jobs offshore (moving jobs offshore may not always be a bad thing, but doing it only because we have silly policy is).

    Of course, most of this goes for business method patents, too. In fact, most ‘software patents’ are filed as business methods, and both cover formulaic or algorithmic processes. You can’t really abolish one without getting rid of the other.

  • http://todd.tv Todd

    There are a couple of other really good reasons why software patents are terrible policy:

    1. The ‘borders’ of a software patent are unclear.

    2. Software patents are detrimental to startups.

    3. Software patents in the U.S. are causing us to ship jobs offshore (moving jobs offshore may not always be a bad thing, but doing it only because we have silly policy is).

    Of course, most of this goes for business method patents, too. In fact, most ‘software patents’ are filed as business methods, and both cover formulaic or algorithmic processes. You can’t really abolish one without getting rid of the other.

  • http://www.kowabunga-dude.com dennis parrott

    In response to Lewis’ reluctance to simply toss software patents into the dumpster (where they damn well belong), I have this to say:

    All that software patents did was to tilt the playing field in the direction of larger entities who have phalanxes of lawyers to file patents on their software navel lint! When s/w patents became (horrid!) law, they severely disadvantaged ALL small players then in the marketplace and ALL players to ever follow in the marketplace because these patents that have been granted for OBVIOUS and PRIOR ART by a corrupt and stupid patent agency has made it such that small players simply cannot compete.

    The NTP patents were bogus as all hell. So are these dingbat database patents. So was Eolas’ patent for what we old assembly language hackers used to call a “wedge”. It is IMPOSSIBLE to write code without standing on the shoulders of those who came, saw and coded before you.

    Software and business processes NEVER should have been allowed to occur. People spoke out against them at the time but Big Business got their way by buying what they wanted from the legislative branch.

    Mark my words well. Software patents are slowly eroding our ability to compete. We are headed to 3rd world status and the slide is slippery.

  • http://www.kowabunga-dude.com dennis parrott

    In response to Lewis’ reluctance to simply toss software patents into the dumpster (where they damn well belong), I have this to say:

    All that software patents did was to tilt the playing field in the direction of larger entities who have phalanxes of lawyers to file patents on their software navel lint! When s/w patents became (horrid!) law, they severely disadvantaged ALL small players then in the marketplace and ALL players to ever follow in the marketplace because these patents that have been granted for OBVIOUS and PRIOR ART by a corrupt and stupid patent agency has made it such that small players simply cannot compete.

    The NTP patents were bogus as all hell. So are these dingbat database patents. So was Eolas’ patent for what we old assembly language hackers used to call a “wedge”. It is IMPOSSIBLE to write code without standing on the shoulders of those who came, saw and coded before you.

    Software and business processes NEVER should have been allowed to occur. People spoke out against them at the time but Big Business got their way by buying what they wanted from the legislative branch.

    Mark my words well. Software patents are slowly eroding our ability to compete. We are headed to 3rd world status and the slide is slippery.

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

    81e31de21f46 Very good http://www.abc-acupuncture.com/baxqorav tramadol

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

    81e31de21f46 Very good http://www.abc-acupuncture.com/baxqorav tramadol

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

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

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

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

Previous post:

Next post: