1. 20

  2. 7

    It seems like if HEVC Advance gets what they want (not sure if RAND/FRAND requirement applies?), then this would likely either just drive people to vp9, or simply keep them using h.264.

    1. 4

      The thing about video patents is that it’s not even a matter of patents that cover particular formats. The MPEG.LA pool has patents that cover, so far, every clean-room codec anyone has tried to invent, including VP9, which was specifically an effort to avoid that.

      But it seems premature to worry about this HEVC group before they actually have any patents. :)

      1. 5

        These particular patents have always confused me. I thought that mathematics couldn’t be patented. How does image processing not fall under the “mathematics” veil? And even if image processing doesn’t constitute mathematics, how does patenting an image processing algorithm not constitute an instance of “patenting an idea rather than an implementation”? I could possibly understand patenting IC designs that efficiently implement encoding/decoding of codecs –though even in that case, it seems that often there is only one design that makes any sense. But the process itself? How does such a patent even get granted?

        1. 10

          Roughly, because patent officers don’t understand the patents they’re granting.

          True, mathematics and software cannot be patented. But “a method and a device” can! So, just rephrase your mathematics and software patent to indicate that it has to be running with a computer. And what software doesn’t! So now what’s patented is the combination of software and mathematics together with this other machine that executes them, so bam, now you have patents of things that were never originally meant to be patented.

          Look what I found this way! A software patent for refusing file transfers if they are too big.

          1. 2

            Yes, this. It’s a reasonable question, but there’s a lengthy case history (Wikipedia) drawing various lines for eligibility, none of which are that.

          2. 2

            I think “can’t patent math” is an oversimplification that misses the larger context and motivation. You don’t want people patenting “F = ma” or other underlying scientific principles.

            Looking at a video codec (or encryption cipher), it’s pretty obvious it’s not a natural property of the world around us.

            1. 4

              But any math is universally true. I don’t really see a difference between “F = ma” and an algorithm with a certain complexity or performance in some particular context. Both are just properties of our universe (or whatever, I’m not a physicist). The problem I see is that it is very hard to make a case like the one I’ve just made without ending up arguing against ALL patents. After all, even a bridge design, one of the classic examples of a “good” patent, can be reduced to a bunch of math that was just waiting to be discovered. Maybe this is why we struggle with this topic so much.

              1. 2

                Yeah. I mean, I tend to feel that very few patents are legitimate, and I don’t have an example to hand of one that I feel is. The whole theory is that they disclose how a thing is done for the betterment of humanity, and get the exclusivity in exchange… but the only ones I can think of that actually patent how they do something, instead of the goal it accomplishes, are the ones patenting something that people were already doing beforehand.

                That is: If it wouldn’t be obvious to a reader how a claim could be accomplished, patent authors leave the details out, and end up getting actually better protections than if they’d disclosed, since then it would have narrower coverage. A system that encourages that is deeply broken.

                Of course, I’m pretty sure there is no set of facts and qualifications that would get that position taken seriously, so it’s just a personal belief. :)

                1. 1

                  I do not believe that IDEA is just a property of the universe, except to the extent that it is part of our universe.

                  Or, in short. You can patent inventions. You can’t patent discoveries. Determining what’s an invention and what’s a discovery is the kind of problem that’s only difficult on Internet forums.

                  (Not to say the patent office doesn’t fuck up, either, but this isn’t meaning of life hard.)

              2. 1

                What others have said, but in addition, it is easy for us to see that software patents and even many hardware patents should be illegitimate, but you have to remember that the general public, pointy haired bosses, judges and even most of the patent office apparently don’t know how computers, networks or software work.

                Hell, a manager I know (Who is very freaking high up) recently declared, “Don’t put any money into the website or [Windows/Mac] desktop builds. Apps are the future, everyone is using apps now”. Meaning he doesn’t know the difference between the desktop clients, mobile apps, the website and how they all interact. He is literally in charge of 10,000+ people, billions of dollars worth of profits/spending/infrastructure and he doesn’t even have a high school understanding of IT.

              3. 2

                Great point regarding video patents. It will certainly be interesting to see how this all “plays” out (pun!). Exciting times! :)

                1. 1

                  Haha, yeah :)

            2. 2

              Still no Daala?

              1. -5