LWN: Comments on "HTML5 video element codec debate reignited"

HTML5 video element codec debate reignited

nix — Sat, 13 Feb 2010 12:29:48 +0000

It's a very sad commentary on the US legal system that it's turned into a
game of 'who will run out of money first' so damn often. Nobody even
*mentions* justice anymore.

HTML5 video element codec debate reignited

robert_s — Sat, 13 Feb 2010 01:29:31 +0000

"A lot of things are patented in all kinds of countries. The question is whether or not these patents can actually be enforced."

When a legal situation becomes this close, it actually boils down to this: if the MPEG-LA decide to drag this backwards and forwards through court, who do you think will run out of money first, MPEG-LA and its industry backers or you / mozilla.org / EFF / etc.?

It won't be the MPEG-LA.

HTML5 video element codec debate reignited

jrincayc — Fri, 12 Feb 2010 03:11:43 +0000

Motion JPEG is widely supported (Gstreamer good, Quicktime and Media Player), and is patent free. However, it requires something like 10+ times the bandwidth that theora does. MPEG-1 is or will be soon patent free (ignoring the MP3 portion), but it is probably at least a little worse than Theora for bandwidth.

HTML5 video element codec debate reignited

marcH — Wed, 10 Feb 2010 14:16:00 +0000

> I'm not sure what you mean by "fix the real problem".

Fix the software patents problem in Europe and use H.264 here. Then not care care about the US and let it deal with its own mess. Not easy but doable.

iPad does in fact support H.264

marcH — Wed, 10 Feb 2010 13:00:32 +0000

> Just like PNG and Jpeg are not mutually exclusive.

OK this is off-topic but... JPEG and PNG serve two very different purposes. The former is for photographs while the latter is for computer images. Use the wrong one and you will get poor quality or poor compression or both.

You probably meant: "just like GIF and PNG are not mutually exclusive".

iPad does in fact support H.264

marcH — Wed, 10 Feb 2010 10:00:58 +0000

Wow, now I realize what this weird "respectively" meant!

I do not think that the use of English makes much difference; this sentence is just too convoluted (in any language).

Otherwise a very interesting summary.

HTML5 video element codec debate reignited

skissane — Wed, 10 Feb 2010 02:07:12 +0000

On a related issue, I would really like the HTML5 standard to require a minimum <audio> and <video> codec, even if its something really old and dumb. There are lots of bit-rate applications (sound in web applications, animations) where something old and dumb is all that's needed, and it would be nice to have some requirement in this area, even if we can't agree on one for the higher-bitrate applications?

For example, an uncompressed or low compression audio format (e.g. some variant of .WAV) is great for adding custom alert noises to a web app, so support for such a format should be mandatory. Similarly, if some old/dumb video format was mandatory, it might be unusable for real video, but still might work perfectly well for animations, presentations, etc.

YouTube Ogg/Theora comparison

dwmw2 — Tue, 09 Feb 2010 12:02:41 +0000

This comparison may be of interest.

In reference to Chris DiBona's comments on WhatWG about Theora and YouTube, it says the following:

"While different files may produce different results, the allegation made on WhatWG was so expansive that I believe a simple comparison can reliably demonstrate its falsehood.
"I do not believe Chris intended to deceive anyone, only that he is a victim of the same outdated and/or simply inaccurate information that has fooled many others.

HTML5 video element codec debate reignited

brettle — Mon, 08 Feb 2010 23:04:28 +0000

Has Microsoft taken any stand either the <video> element or Theora? If they added support <video>+Theora they might be able to:

1. Help defeat Adobe Flash.
2. Get some cred from the open source community.
3. Save a little money on MPEG-LA licensing fees in the long run by making H.264 support a downloadable option.

Suing individual users of H.264

Trelane — Mon, 08 Feb 2010 21:10:36 +0000

If you used it, you would be infringing.

If you used an encoder/decoder that you wrote, you'd be infringing twice (see the USPTO link above; the act of making a patented idea is prohibited, as is using an implementation without a license).

Suing individual users of H.264

roc — Mon, 08 Feb 2010 20:50:38 +0000

> In the US, my understanding (IANAL) is that practicing a patent does not
> require a license, only importing or distributing.

Incorrect. Practicing requires a license.

> Certainly, if I were to write my own H.264 decoder, I am not infringing.

Incorrect. If you used it, you would be infringing.

Suing individual users of H.264

Trelane — Mon, 08 Feb 2010 04:01:33 +0000

"In the US, my understanding (IANAL) is that practicing a patent does not require a license, only importing or distributing."

The USPTO Sez (http://www.uspto.gov/web/offices/pac/doc/general/#patent)

The right conferred by the patent grant is, in the language of the statute and of the grant itself, the right to exclude others from making, using, offering for sale, or selling the invention in the United States or importing the invention into the United States. What is granted is not the right to make, use, offer for sale, sell or import, but the right to exclude others from making, using, offering for sale, selling or importing the invention. Once a patent is issued, the patentee must enforce the patent without aid of the USPTO.

Relevant portions bolded and italicized by me. (See also the the MPEG-LA's response to my query in the other article comments.

Certainly, if I were to write my own H.264 decoder, I am not infringing.

Yes, you would. That would be the non-bolded "making" above, which precedes the bolded "using".

Now, if I were to distribute a software decoder, that would require a license under current case law.

Perhaps under current case law (IANAL), but not under a strict reading of the USPTO's talk (as long as it's not for sale, sold, or brought into the US).

Suing individual users of H.264

jhhaller — Mon, 08 Feb 2010 02:07:22 +0000

Suing end-users would be dependent on the particular domestic law. In the US, my understanding (IANAL) is that practicing a patent does not require a license, only importing or distributing. Certainly, if I were to write my own H.264 decoder, I am not infringing. Now, if I were to distribute a software decoder, that would require a license under current case law. Surprisingly, there has been no effort as of yet to stop US-based mirrors of distributions containing FFMPEG. Of course, this case law is under review by the US Supreme Court, so this uncertainty could be inhibiting bringing of lawsuits, as they could make such lawsuits easier or harder. This leaves the typical method of getting H.264 software, importing it into the US. The International Trade Commission can block import of products containing patented technology, although how that would play out against multiple, independent importers could be interesting.

iPad does in fact support H.264

roc — Sun, 07 Feb 2010 22:08:07 +0000

> The RIAA is not a patent pool, this is a straw man.

The point is that "<Large Corporate Entity> would never sue end users, because they have no money" is an argument that has failed in the past.

>> Plus, let me remind you that your guess about the intention of the
>> MPEG-LA to "not require" licenses for free software proved to be
>> completely wrong.
> I never said any such thing.

http://lwn.net/Articles/371439/
> You don't need a license from the MPEG-LA (and neither does FFmpeg)
> because you don't qualify for requiring one:

iPad does in fact support H.264

man_ls — Sun, 07 Feb 2010 12:01:47 +0000

While in theory they could do that in the future, it's about as likely as getting killed by a meteor hit.

Meteors scare me, man! Just the thought of them makes me afraid. Often fear (not likelihood) is enough to prevent people from doing things.

Submarine patent threat still valid?

giraffedata — Fri, 05 Feb 2010 21:38:45 +0000

But we're not talking about nondisclosed patents here. A submarine patent is a patent that doesn't exist yet.

So you'd have to go further with estoppel and find a duty of an inventor to meet some standard of effort in getting the patent office to grant the patent quickly. It would be rather difficult to prove that the patent didn't issue sooner because the inventor wanted people to use the invention royalty-free during that time.

HTML5 video element codec debate reignited

Simetrical — Fri, 05 Feb 2010 19:36:11 +0000

Not true according to this blog post by the implementer:

We believe that the web platform must be built on open standards and will therefore continue to support the Ogg formats: the Vorbis audio codec and the Theora video codec. These, in addition to plain WAVE PCM audio, are our "core codecs" which we will support on all desktop platforms. . . .
For this release . . . we have adopted the GStreamer media framework as an extra layer between the browser core and the raw decoding. Among other things, this allows processing to take place in a separate thread, which has improved responsiveness and audio quality.
For platforms where GStreamer is natively available, we are simply using the system-installed version. Thus, if you are using Linux or FreeBSD, make sure to install at least the GStreamer "base" and "good" plugins, otherwise <video> won't work at all. . . . Having done this, Opera will be able to play anything that GStreamer can handle . . . We hope you have fun playing with this, but stick to Ogg for anything serious that should cross-platform and cross-browser.
On Windows we have made a minimal GStreamer configuration which keeps only the features necessary to decode the above mentioned core codecs. . . .

On Windows (and presumably on Mac when support for that is added), Opera supports only Theora for video. On Linux/BSD, it uses system GStreamer libraries and will support anything they do.

HTML5 video element codec debate reignited

Simetrical — Fri, 05 Feb 2010 19:31:47 +0000

It really is simple cause-and-effect.

1) The spec used to require Theora support. Apple said they wouldn't
support Theora, even though the spec required it at the time. Therefore,
the lack of a requirement cannot be why Apple doesn't support Theora.

2) When the requirement was removed from the spec, the reason given by the
editor was explicitly that Apple did not support Theora. Therefore, this
was the cause, unless you want to accuse Ian of lying. (Do you?)

Apple never asked for the requirement to be removed, as far as I know. They
just said they would ignore it if it wasn't, so Ian made the decision to
remove it.

HTML5 video element codec debate reignited

n8willis — Fri, 05 Feb 2010 19:01:18 +0000

The cause-and-effect given here is exactly backwards. HTML5 does not mandate any format because the war is underway. The editor, Ian Hickson, is not willing to add anything to the standard if a major player refuses to implement it, because then it's not a standard, it's a work of fiction. Apple refuses to implement Theora support, and Mozilla refuses to implement H.264 support, regardless of what the spec says, so it would be pointless to try mandating either it would just make the spec less useful to anyone who expects it to be consistently implemented as written.

I don't agree. I don't think it's cause-and-effect, for starters -- these two activities are intertwined and simultaneous. The H.264 stakeholders know and accept that W3C will not include a royalty-collecting format in a standard; they are not seeking to have H.264 be declared part of the standard, they are trying to prevent any competing format from becoming part of the standard, because that decision would seriously hurt them in the ongoing de-facto war for dominance.

Nate

HTML5 video element codec debate reignited

gmaxwell — Fri, 05 Feb 2010 19:00:53 +0000

Yes. As a format H.264 has a lot of attractive things in it and among those tools in H.264's collection is basically a superset of Theora. If your technical comparison points are only quality-per-bit and not decoder computational complexity, decoder implementation complexity, licensing, or... "Assume an ideal spherical encoder in a frictionless market" H.264 has more signal processing tools in the decoder, so it will win that kind of comparison.

Though it's important to note that what we can actually compare the quality-per-bitrate of is encoders, not formats.

Quoting Timothy Terriberry,

"There are _so_ many things you can do wrong in an encoder that do much more harm to quality than a clever optimization scheme or a complicated, patented scheme can improve it. Most encoders do some or all of them, and the original VP3 encoder was no exception. With Thusnelda we're doing things a lot smarter than we used to be, and that will only get better."

Libtheora doesn't beat the quality-per-bit of best H.264 encoders (such as x264), and won't if they also keep improving, but it does very well against some very popular ones. So thats another layer to question of 'good enough for the proposed uses and whether or not other tradeoffs make Theora preferable': "You're already using an inefficient H.264 encoder, Theora isn't worse than that. Why not use Theora and help contribute to a world without format royalties?"

Sun OMS

giraffedata — Fri, 05 Feb 2010 16:57:50 +0000

It's worked well for MPEG— they claim responsibility for over $66 from every person on earth. (Also, checkout their development man-hour estimates. It's rather daunting to go up against that)

This is highly misleading. I checked out the link (which is just the visual aids for a presentation I didn't see, so take even this clarification with a grain of salt).

The $66 is for all products that use MPEG-2 in some way. It would be a mistake to say that MPEG-2 is responsible for all that value. And it's pretty arbitrary, because e.g. it includes the cost the screen in my TV, but not of the couch in my TV room.

The man-hour estimate given is one million people working for 15 years. That's obviously not people working on MPEG-2. My guess is it's people working on developing and delivering products that use MPEG-2 in some way, like the guy who created the injection molds for my TV's case.

HTML5 video element codec debate reignited

DonDiego — Fri, 05 Feb 2010 13:12:39 +0000

Another correction:

> H.264 supporters claim that Theora's quality-per-bitrate performance is behind H.264's

This is not a claim but a technical fact that not even the Theora developers dispute. The question is whether or not Theora is good enough for the proposed uses and whether or not other tradeoffs make Theora preferable over H.264.

HTML5 video element codec debate reignited

DonDiego — Fri, 05 Feb 2010 13:02:46 +0000

* Opera also uses system codecs, so H.264 support is available if you have a codec installed on the system.

iPad does in fact support H.264

DonDiego — Fri, 05 Feb 2010 11:34:44 +0000

I would like to politely and constructively suggest to refrain from such elaborate sentence construction. Many of your readers are not native speakers and will have trouble grokking what you are trying to say. I myself had to reread the sentence although I think I can claim with some confidence that, for a non-native speaker, I have an above-average command of the English language.

HTML5 video element codec debate reignited

DonDiego — Fri, 05 Feb 2010 10:58:51 +0000

> > The whole discussion about H.264 vs. Theora and the MPEG LA really
> > only concerns a small part of the world.

> Yeah, just the USA, Europe, and parts of Asia.
> http://weblogs.mozillazine.org/bz/archives/020400.html

A lot of things are patented in all kinds of countries. The question is whether or not these patents can actually be enforced. I'm looking forward to you posting proof of enforcement outside of the USA.

iPad does in fact support H.264

DonDiego — Fri, 05 Feb 2010 10:50:36 +0000

> There's no way to tell. Five years ago you could have said the same thing about the RIAA suing individual file-sharers, but then they started doing it.

The RIAA is not a patent pool, this is a straw man.

> Plus, let me remind you that your guess about the intention of the MPEG-LA to "not require" licenses for free software proved to be completely wrong. Betting on such guesses, no matter who makes them, would be foolish.

I never said any such thing. The MPEG LA patent license question rests on the depth of your pockets, not on the type of software you use.

HTML5 video element codec debate reignited

wookey — Fri, 05 Feb 2010 04:43:38 +0000

Hey that's great - it actually works in my browser! As an 'optimist' who has been persevering with gnash for the last year or two, I dont see much video: currently youtube flash doesn't work and youtube html5 h264 doesn't work either. BBC iplayer doesn't work, and most video embedding on random other sites doesn't work either. Sadly sites very rearely provide a nice 'download the video' link anymore (which works just fine). I normally only ever get to see things via get_iplayer, clive, youtube-dl and the like.

So, yes vfe looks good from here.

Submarine patent threat still valid?

AndreE — Thu, 04 Feb 2010 22:24:23 +0000

Well the submarine patent argument is a bit of a strawman because it is equally valid for PATENTED software as well.

Look on MPEG-LA page where they state that licensing the patent from them doesn't guarentee against litigation by other parties.

Also cross reference the Alcatel-Lucent v. Microsoft case where AL sued Microsoft for mp3 patent infringement despite Microsoft already paying Thompson license fees.

Software patents are such a joke because no one is absolutely sure who own the patents to what exactly. It looks very much like a cartel managed by "legitimate businessmen" if you know what I mean

HTML5 video element codec debate reignited

dr@jones.dk — Thu, 04 Feb 2010 22:14:48 +0000

Regarding non-java alternatives: Please consider mentioning http://camendesign.com/code/video_for_everybody

iPad does in fact support H.264

roc — Thu, 04 Feb 2010 21:09:38 +0000

> The MPEG LA has never sued an end-user to date. While in theory they could
> do that in the future, it's about as likely as getting killed by a meteor
> hit.

There's no way to tell. Five years ago you could have said the same thing about the RIAA suing individual file-sharers, but then they started doing it.

Fundamentally it's a bad idea to bet on the MPEG-LA being lenient forever. Their job is to bring in licensing revenue for patent holders. At some point in the future, if suing users is a convenient way to scare people, or to destroy free competition to products that actually generate license revenue, there's no reason to believe they won't do it. Keep in mind that the optimal revenue-generation strategy changes over time: it pays to be lenient at first, to get the format maximally entrenched, and then you can squeeze the licensees for all you can get (modulo contractual restrictions). History is instructive.
http://www.0xdeadbeef.com/weblog/2010/01/html5-video-and-...

Plus, let me remind you that your guess about the intention of the MPEG-LA to "not require" licenses for free software proved to be completely wrong. Betting on such guesses, no matter who makes them, would be foolish.

iPad does in fact support H.264

roc — Thu, 04 Feb 2010 20:59:48 +0000

> Who uses Vista? ;-p

Some people. I was just correcting an error in the poster's facts.

> So what are the statistics?

I believe currently around 60% of our users are on WinXP. Not sure how many are on Vista. The fraction of those people who have installed a DirectShow H.264 code is probably tiny, so we can expect that somewhat less than 40% of our users have an H.264 codec on their system.

HTML5 video element codec debate reignited

kov — Thu, 04 Feb 2010 19:24:25 +0000

Using a Java applet is not a good idea. Java is not well supported in many platforms, and overall does not work well, in my experience.

About Youtube's move, it was funny reading their announcement; they said they were moving in that direction as an answer to the most voted feature "idea", but totally ignored the 'open formats' part of the request they claimed to be responding to.

HTML5 video element codec debate reignited

anton — Thu, 04 Feb 2010 19:14:18 +0000

[...] using H.264 for Youtube is beneficial for Google, since it potentially gives Chrome more market share [...]

The other side of the medal is that a competitor of Youtube who supports formats supported by Firefox can increase their market share at the cost of Youtube.

Mike Melanson on HTML 5 video

DonDiego — Thu, 04 Feb 2010 17:41:13 +0000

I cannot help but post Mike Melanson's take on HTML 5 Video, the codecs used therein and how it competes with Flash. Humorous quote:

Another aspect I have to appreciate about the debate surrounding HTML5 video is the way that it brings out the positive spirit in people. Online discussions are normally overwhelmingly negative. But advocates of the HTML5/Xiph approach truly believe this could all work out: If Apple decides to adopt the Xiph stack, and if some benevolent hardware company would churn out custom ASICs for decoding Xiph codecs, and if those ASICs were adopted in next quarters array of mobile computing devices and netbooks, and if Google transcodes their zillobytes of YouTube videos to the Xiph stack, and if Google throws the switch and forces the 60% of IE-using stragglers to either change browsers or go without YouTube, and if Google thereby forgoes many opportunities to monetize their videos, then absolutely! HTML5 video could totally unseat Flash video.

Ironically Mike is both the main person working on the Linux port of Adobe Flash and the original author of the VP3 spec on which Theora was based. He always had a weak spot in his heart for fringe multimedia formats, but he surely had no idea what kind of genie he was letting out of the bottle there...

iPad does in fact support H.264

quintesse — Thu, 04 Feb 2010 16:48:35 +0000

RedHat doesn't distribute ffmpeg exactly for that reason I guess.

iPad does in fact support H.264

blitzkrieg3 — Thu, 04 Feb 2010 16:44:39 +0000

> The MPEG LA has never sued an end-user to date.

They would not sue an end user. End users have no money. They would sue a Red Hat, or a SUSE for giving you access to ffmpeg package.

iPad does in fact support H.264

DonDiego — Thu, 04 Feb 2010 15:56:31 +0000

> In many countries, the MPEG-LA could sue you (and win) for using *or distributing* ffmpeg's H.264 decoder. Is that the kind of regime you want to live under?

The MPEG LA has never sued an end-user to date. While in theory they could do that in the future, it's about as likely as getting killed by a meteor hit. Microsoft could also sue you for using Linux. You never know, they do have patents that cover it.

If you are afraid of distributing an H.264 decoder, fine, leave it to others and use what is available on the system.

> > As far as other platforms XP does not support H.264 out of the box, but Vista and Windows 7 does.

> Vista doesn't.

Who uses Vista? ;-p

Seriously, how big a slice of your userbase uses Vista? For the FFmpeg and MPlayer websites I see less than 20% of the Windows users on Vista, most are still on XP and likely to skip Vista on their upgrade path.

So what are the statistics? How many of your users do not have access to a system H.264 decoder and how many will not in five years time?

HTML5 video element codec debate reignited

Simetrical — Thu, 04 Feb 2010 15:49:26 +0000

This article perpetuates a common and insidious myth about the video format war:

The HTML 5 standard does not mandate that support be included for any particular format in order to qualify as compliant, however, so a public war is underway between format proponents for de-facto dominance.

The cause-and-effect given here is exactly backwards. HTML5 does not mandate any format because the war is underway. The editor, Ian Hickson, is not willing to add anything to the standard if a major player refuses to implement it, because then it's not a standard, it's a work of fiction. Apple refuses to implement Theora support, and Mozilla refuses to implement H.264 support, regardless of what the spec says, so it would be pointless to try mandating either – it would just make the spec less useful to anyone who expects it to be consistently implemented as written.

Your own duplication of Ian's mailing list post says this (emphasis added:

. . . I have reluctantly come to the conclusion that there is no suitable codec that all vendors are willing to implement and ship. . . . I have therefore removed the two subsections in the HTML5 spec in which codecs would have been required . . .

Please fix this error to avoid confusing anyone further.

Some other minor errors:

MPEG-LA is not "pushing for adoption of its H.264 format" on the web, that I know of. If anyone is, it's Apple and Google. But I wouldn't even saying they're "pushing" for H.264 adoption, they're just using it themselves (in Google's case) and not supporting Theora (in Apple's case).
Safari supports anything that QuickTime does, as I understand it. In particular, it does support Theora if you install the right codec.
Opera doesn't support <video> at all in their current releases. The 9.50 development versions support only Theora on most platforms, but AFAIK, they use system GStreamer on Linux and so will support H.264 there in many cases.
You can enable H.264 support in Chromium/WebKit if you like, obviously. The Chromium PPA for Ubuntu supports it if you have non-free codecs installed.

Submarine patent threat still valid?

gmaxwell — Thu, 04 Feb 2010 15:17:46 +0000

Right, patents are not "use it or lose it" like trademarks. However, estoppel is a general principle of law and some have proposed more aggressive use of estoppel as a general tool to deal with the risk of non-disclosed patents gumming up the creation and adoption of standards. (I highly recommend that paper: it provides a good view 'down the rabbit hole' of the current mess that exists between patents and standards)

iPad does in fact support H.264

n8willis — Thu, 04 Feb 2010 14:35:30 +0000

I love clauses; the more, the merrier -- and for that I apologize to no one.

Nate