H.264 vs. Theora — Fightin' Words About Patentability 421
An anonymous reader writes "Thom Holwerda from OS News has penned a rebuttal to claims from Daring Fireball's John Gruber that Theora is a greater patent risk than H.264. Holwerda writes, 'And so the H264/Theora debate concerning HTML5 video continues. The most recent entry into the discussion comes from John Gruber, who argues that Theora is more in danger of patent litigation than H264. He's wrong, and here's why.'"
When h.264 isn't h.264. (Score:1, Informative)
"Today, people have data in H.264 format. A lot of data. "
And a lot of it totally useless. You see-- full h.264 is so gobsmackingly processor intensive that even a high end multi-core desktop with hardware assist can have a decently hard time of keeping up with it. More importantly some of the features in h.264 are incredibly hostile to fast hardware implementation: For example, the arithmetic coder is purely serial and can't be made to operate in parallel, so the only way to make it fast is to run it at a high clock speed and high clock speed means high power.
Because of this many "h.264" devices, like the iphone, only support an exceptionally watered down profile called baseline. Most of the technical features h264 has over theora are gone in baseline, b-frames and multiple references, the arithmetic coder, adaptive quantization, 8x8 transform, etc.
So when you have a "h.264" file you really have no clue when it will play... and if its a "well encoded" file it probably will not play in many places.
Re:Patent risks (Score:3, Informative)
It stretches the definition of 'invention' to claim that algorithms are inventions.
Second, we don't need any argument to extend patents to algorithms. They already are. If you want to abolish software patents, however, you should make a good case for that move.
They aren't patentable in most of the world, nor were they even here for the period of most explosive growth of software. Software patents are a bit of an anomaly, and their presence still needs justification.
Re:Patent risks (Score:3, Informative)
Re:First Post (Score:1, Informative)
Flash uses h264 for video compression since Flash 9.
Re:First Post (Score:3, Informative)
And that's what it's doing right now. HTML5 itself is codec-agnostic.
The problem is that it doesn't matter what codecs are supported, it matters what codecs people actually use to post content online. If you can't legally support H.264, and people want to post H.264 videos, you're screwed.
Now, can browsers support the system codecs? Yes, and I've heard claims that on Unix, Opera actually does. Problem solved, anything that you have a GStreamer codec for can be played in Opera using <video>.
However, part of the point of <video> was that it would get rid of having to download codecs and media players and such.
Re:Patent risks (Score:3, Informative)
Oh ye of little imagination. The algebraic expression you provided is a parameterized program with unknown variables x, a, b, c. You have provided exactly a simple program which can be evaluated, once the variables are provided, by the rules of algebra, which form the instructions of the machine that can execute this program.
See the Curry-Howard correspondence for a formal proof of this. Types are logical propositions, and programs are proofs. It is a mathematical fact that programs are math, that the instructions of a machine correspond to the axioms of a formal/mathematical system, and that "sets/sequences of instructions" correspond exactly to mathematical proofs.
Re:First Post (Score:2, Informative)
That way it'd be the same as the change from GIF to PNG all those years ago, where those who want to use GIF could, and those who needed / wanted the free option (which was also superior) could use it without killing support for the other.
IE didn't support transparent PNG files until IE7, 2006 (yes... 2006). I hope I have answered your question.