Boycott HTML5?

HTML5 and H.264: Video wants to be free

By Rick Jelliffe
February 18, 2010 | Comments: 3

Mozilla's Robert O'Callahan has had a great series of blogs (the comments are good too) over the last month:

Robert gave a talk at the recent Linux conference The importance of open video on the Web.

The money quote (his bolding):

Youtube and Vimeo have started offering video playback using the HTML5 <video> element. That is good news for free software, since it means you don't need a closed-source Flash player to play the video [1]. However, they only offer video in H.264 format, and that is not good news for free software. A lot of people have noticed that Firefox doesn't support H.264, and apparently many people don't understand why, or know what the problems are with H.264. ...

The basic problem is simple: H.264 is encumbered by patents whose licensing is actively pursued by the MPEG-LA. If you distribute H.264 codecs in a jurisdiction where software patents are enforceable, and you haven't paid the MPEG-LA for a patent license, you are at risk of being sued.

(Christopher Blizzard has a good post on this too.)

Made me laugh, anyway

In Robert's blog he also mentions a talk by Samba's Jeremy Allison:

He made the point (which I think is too often overlooked) that which company one works for is almost always an individual moral choice and we should hold people accountable for it ...

I am not sure what "hold accountable" means. It is probably a euphemism for being nasty and self-righteous. Yes, this is just what the FOSS movement needs to get ahead: increased toxicity.

Allison is then reported:

Jeremy suggested that Microsoft will promote "RAND" standards --- standards covered by patents whose licenses would require a "Reasonable And Non-Discriminatory" fee, which sound good except that for free software, any non-zero fee is a show-stopper.

But, hold on, isn't that just what YouTube is actually doing by using H.264 for their codec? And doesn't Google actually own YouTube? And doesn't Jeremy Allison actually work for Google?

So according to his (reported) comment, shouldn't we be holding him accountable for working for a company which is actually doing what he condemns his pet enemy for being about to do in his dire prediction? I must be missing something...

Piddly hypocricy and party spirit aside, I do really agree with (what I take as) the thrust of what O'Callahan (and Allison too, probably) are on about.

For an Open Base

I think genuinely Open standards (especially with public or national interest) need to have at least one way to do each feature that allows FOSS development. Both free and open source must not be excluded.

And preferably that way should be available as a base-line with other FOSS-unfriendly formats and codecs available by negotiation: optional extras for extra quality or features. I support plurality and an Open base. (For an application of the same principle w.r.t. XML markup, see Safe Plurality, with a further wrinkle at Supporting Degradation.)

I understand that HTML5's <video> element is agnostic as far as video format or codec is concerned: that is normal good layering. But in the current environment, it looks like being a mistake for HTML5 to, in effect, enshrine single formats and codecs: it seems reasonable to expect that it will be the non-FOSS new media/advertising companies who will determine the winner. The YouTubes not the Microsofts.

I have written before about how important it is for corporations to have control of the API (for example, The Frog Race about the benefits of competitive standards rather than monocultures.) It is not a conspiracy, just something that requires counter-action.

IMHO, the reasons that Linux/Solaris/BSD has so far failed on the desktop are three:

  • Microsoft's bundling strategies

  • the concept of a Standard Operating Environment (and here), which frames procurement questions in a way that encourages concentration

  • the relative lack of media and gaming capabilities, in particularly caused by MPEG licensing difficulties with FOSS

The message I get from Robert O'Callahan's blogs is that this third reason is in strong danger of being perpetuated, perversely by Google. I though they were supposed to be Microsoft's great enemy or something. (By the way, the same comment can be made about the second reason and IBM: by emphasizing the SOE, even if it is to sell Lotus Notes solutions, IBM paradoxically perpetuates a meme that helps maintain Window's market dominance.)

The only game in town

There is only one game in town at the moment where this could be addressed that I know of: the HTML5 effort.

It looks to me that HTML5 needs to have a better story on allowing adequate plurality for <video> that does not marginalize FOSS (and boutique) developers, especially smaller poorer ones, sole players and hobbyists. This is the danger of standards being made by groups dominated by corporate oligopolists: they give themselves the delusion that they can speak for their users when they really just speak for themselves.

But we must not have the situation where chunks of HTML get, in effect, proprietorized: if there is a video on a webpage, I should be able to view it even in low-fidelity, from my FOSS software. Video wants to be free.

[Readers: please discuss details of H.264 licensing elsewhere, such as O'Callahan's blog, since it is outside my expertise. But corrections to errors here are welcome of course, and discussion of the FOSS problem.]

You might also be interested in:


It's been somewhat lost in the H.264 vs Theora quality arguments on the web recently but the W3C, as a rule, won't promote patented standards. Therefore there was never any chance of H.264 being specifically referenced in the HTML5 spec and no-one has ever even suggested it because they know it would be rejected.

Theora was originally suggested in the spec as a baseline interoperable format that could be implemented by everyone but not in any way mandated as the only codec or any suggestion that it was the "best" codec except for being the best of the group that lack patent royalties (e.g. Motion-JPEG was seriously suggested as a potential fallback video format, which is just a flickbook of JPEGs) . This very sensible approach is, I gather, basically what you are suggesting here.

This was scuppered by Apple and Nokia. (Apple and Nokia are of course MPEG patent holders, normally this would seem paranoid to point out apart from Nokia only revealing this info long after they'd made a very peculiar stand against Theora, calling it "proprietary" and claiming it was impossible to add DRM).

Apple said Theora wasn't good enough (which makes no real sense for a baseline format), they said they'd already produced millions of devices that didn't support it in hardware (which doesn't matter as Theora is lightweight enough to play with updated software on all of them) and they said the patent landscape was uncertain so they'd be taking on extra risk, and they'd already committed to H.264 (which implies there are people who hold patents that apply to a simple video codec like Theora, but are miraculously avoided by the far more complex H.264 codec. Basically this plays of people's misunderstanding that believes if you pay MPEG-LA for their patents then no-one else can sue you).

The editor of HTML5 took out the recommendation to use Theora, which had previously been in the spec. His reasoning was that Apple was going to ignore the spec either way, and he didn't want to put something into the spec that was blatantly not going to be true as it undermines the process and he didn't feel that writing it in the spec had any power to change Apple's mind.

Also, on the working for Google is bad if they push patented standards issue:

The great white hope of the codec debate is for Google to release the codecs they just bought from On2 in a royalty free manner. They certainly hinted as much in their announcement of the deal and the shareholders approved it just this week.

Whether this will convince Apple to support this codec or the editor of the HTML5 spec (who is a Google employee) to mandate it in the standard is an interesting question.

I'm hopeful they contribute their newly purchased patents, code and manpower towards an alternative RAND-Z video codec standard perhaps under the IETF rather than just release them as-is. Either way it'll be disruptive to the patent pool approach to video standards.

Dash: Any of these companies could be the great white hope: I don't know why any company that seriously wanted to dent MS' desktop domination (especially by levering on FOSS) would even think of perpetuating the current system. If the all rules of the game don't change, why should the current outcome change?

It would be great if Google was able to come up with some smarter approach to all this. If they commit to only support parts of H.264 that can are unencumbered or RAND-Z-able, for example. (But, yes, Ian Hickson's quixotic approach may be a real spoiler for change here.)

As far as Google being bad, of course I (if it is not clear) think any simplistic summary of a company into a moral boolean is pretty dumb. Just as Torvalds' quote "I may make jokes about Microsoft at times, but at the same time, I think the Microsoft hatred is a disease." hatred of Google, IBM, WHATWG, W3C, etc etc would also be a disease.

News Topics

Recommended for You

Got a Question?