Wikipedia's Assault On Patent-Encumbered Codecs

An anonymous reader writes "The Open Video Alliance is launching a campaign today called Let's Get Video on Wikipedia, asking people to create and post videos to Wikipedia articles. (Good, encyclopedia-style videos only!) Because all video must be in patent-free codecs (theora for now), this will make Wikipedia by far the most likely site for an average internet user to have a truly free and open video experience. The campaign seeks to 'strike a blow for freedom' against a wave of h.264 adoption in otherwise open HTML5 video implementations."

Comments

HTML5 Video

[sopssa]

It's all nice and all, but if open video technology really wants to win, they have to be technically better. There is no other way.

However it's nice to see Open Video Alliance trying to partner with Wikipedia. In addition to being technically better, that's another aspect you need to take care of. You need to make sure websites, TV, phone, computer and so on manufacturers support your technology. You have to work with them to get it supported - not just put it out there and hope it catches up because its "open", because that's not going to happen. Personally I would also hate to see technically inferior solution being used, as it would eat huge amount of bandwidth. Theora just isn't on the same table with H.264 for Internet video. Theora is based on VP5 from On2 and now that Google acquired them, they're going at VP8.

As far as having a single standard for HTML5 video goes, Theora lost. H.264 is and has been already everywhere and on every device. I also suspect majority of sites will use H.264, as that's what is being used with Flash already.

However, what I see happening (and hope) is HTML5 Video tag being released without requiring support for a single codec, just like img tag is. Then browsers can either implement their own support, use third party tool like gstreamer (like Opera does) or just depend on OS (what I suspect IE and Safari will do). Firefox is still having their ideological problems, but I'm pretty sure they will start using gstreamer too.

What I'm more worried about is that I cannot watch Wikipedia videos with any other device than my PC. Want to see a video clip of a place you're traveling on your phone? Not possible. Want to see videos from Wikipedia with your PS3/360? Not possible. It will create some serious problems, and I don't think Wikipedia is big enough to push the change alone.

Re:HTML5 Video

[Anonymous Coward]

Open video bitstream formats, like Theora, are simply not capable of being better than H.264 (yet). The best bet in that regard is Dirac by the BBC, but even that does not fare too well against H.264 as encoded by x264.

However, open video formats simply do not need to be better than the proprietary formats, they simply need to be "good-enough" and be ubiquitous on the web, and pretty soon all browsers (except IE, probably) will support them out of the box. Wikipedia going with theora is a good move in that direction.

Re:HTML5 Video

[Xtifr]

It's all nice and all, but if open video technology really wants to win, they have to be technically better. There is no other way.

Why? Closed formats don't seem to operate under that constraint. In fact, technical qualities seem to be a non-issue as far as success goes in general. The backing of big players seems to be what counts, and that's exactly what we have here. Whether Wikipedia + Firefox + RedHat + other open players is big enough remains to be seen (and I admit I have my doubts), but if "technically better" becomes an issue, I think it'll be the first time ever.

What I'm more worried about is that I cannot watch Wikipedia videos with any other device than my PC

Ah, now your real concern appears, I suspect. If Theora starts to get momentum, it'll appear on phones and similar devices quickly enough. My phone already supports Ogg Vorbis. (It may even support Theora; I haven't tried.) If yours doesn't, then perhaps you went with the wrong vendor. I didn't look for Vorbis support for my phone, but I did look for openness; if that wasn't a factor in your choice of phone, then my sympathy for you is nil. Especially if you want to connect with Wikipedia, whose commitment to openness is legendary.

If you want Wikipedia to go with your proprietary, encumbered format(s), your best be is to lobby the patent holders to donate the patents to the public domain. Good luck with that. :)

Re:HTML5 Video

[Anonymous Coward]

What I'm more worried about is that I cannot watch Wikipedia videos with any other device than my PC. Want to see a video clip of a place you're traveling on your phone? Not possible. Want to see videos from Wikipedia with your PS3/360? Not possible. It will create some serious problems, and I don't think Wikipedia is big enough to push the change alone.

In general I find the "must have hardware support now" argument a bit short sighted. By that reasoning there would never be any change in video codecs. In any case, the PS3 and 360 even combined represent a very small percentage of internet connected devices. And the 360's larger problem is not having a web browser so Wikipedia video would be streamed from your PC anyway and if needs must you can transcode on the fly.

As mobile phones go, my Nokia N900 plays Theora. It also runs Firefox. Fennec is on Maemo 5 (the N900's OS) and will soon be available for Android, Windows Mobile, and future MeeGo devices. Millions of devices in the field already have the capability to play Ogg Theora and it will only become more trivial to do so with Firefox releases for those platforms.

Re:HTML5 Video

[Lunix Nutcase]

But why should the average user be quite in some stupid ideological fight when they are never going to be paying for the H.264 royalties that Microsoft, Apple and Google will be shelling out to include H.264 support in their browser?

Re:HTML5 Video

[DragonWriter]

It's all nice and all, but if open video technology really wants to win, they have to be technically better. There is no other way.

For any definition of "technically better" where this is not a vacuous tautology (that is, any definition other than "technically better means whatever ends up winning"), this isn't true: solutions that aren't "technically better" by almost any definition you choose will win all the time, because the business model behind selling them allows them to be sold cheaper (even if they aren't any cheaper to produce), because they are imposed by market-dominant players, or for all kinds of other reasons beside technical superiority.

Compatibilty of patent-unencumbered formats with a venue like Wikipedia would be exactly that kind of non-technical factor. (As would, on the other side, the competitive advantage that those who co-own the patents see in the dominance of patent-encumbered formats that they are part of the controlling syndicate for.)

Re:HTML5 Video

[AliasMarlowe]

It's all nice and all, but if open video technology really wants to win, they have to be technically better. There is no other way.

Try telling that to Microsoft!
Recall that Windows did not become the de facto standard OS by being better - it was definitely not better than the alternatives in the period in which it became dominant. So there is another way: gain sufficient market share through fair means or foul, and you can win. Whether wikipedia would count as critical mass or not is an open question, but if they were sufficiently bloodyminded, then whichever codec they chose to standardize on would ipso facto become a necessary codec, even if it were not used widely elsewhere.

Re:HTML5 Video

[wealthychef]

The cost is still paid by the average user, it's just tacked onto the cost of the O/S or whatever you buy from Apple, MS, etc. As for "why", various consumers value different things. Some value cheap, some value fast, some value open source, some value high quality, etc. Ideally, customers who want choice can get it with a plugin, and the rest will get it easily without a plugin. But there will always be this creative edge. Most people will just say with safe, reasonably fast and easy.

Re:HTML5 Video

[Anonymous Coward]

If they don't offer hardware acceleration in a mobile device, your begging for a huge battery drain. The point isn't only that they can play them, but that they are well supported by the device and that the impact isn't overly detrimental.

Re:HTML5 Video

[slimjim8094]

To be fair, the format is entirely open, but patent encumbered. Nobody would argue that MP3 is a closed format, for example.

IOW the only challenges are legal challenges (regarding software patents and royalties). They're not proprietary at all.

Re:HTML5 Video

[sopssa]

The cost is still paid by the average user, it's just tacked onto the cost of the O/S or whatever you buy from Apple, MS, etc

Assuming the latest amount of 1.9 billion internet users (and not even accounting those not using internet), the $5 million cap per license, and Windows market share of 98%:

$0.002 per user.

I just don't see so many people caring.

Re:HTML5 Video

[moderatorrater]

It's all nice and all, but if open video technology really wants to win, they have to be technically better. There is no other way.

Why? Closed formats don't seem to operate under that constraint.

Because closed formats have a company or companies willing to push it for reasons other than technical superiority. Open source relies on a lot of people getting excited about something and pushing it in a more organic way, and for that to happen in a big way then the thing they're pushing has to be technically superior. Linux has gained momentum in its areas by being superior for developers and sysadmins who know what they're doing. Firefox gained momentum the same way. I can't think of an open source product that gained mainstream popularity without being technically superior.

I didn't look for Vorbis support for my phone, but I did look for openness; if that wasn't a factor in your choice of phone, then my sympathy for you is nil

So, only people who spend their valuable time and money getting an open phone instead of the iPhone are worthy of consideration in this debate? Like it or not, the iPhone's dominance isn't because of any media blitz or cult of Apple, it's because it came out in a market where it was by far the best choice and is still superior to any other smartphone I've seen.

So, if you want to prioritize openness in your purchasing, that's fine. But this is about Wikipedia trying to influence the culture as a whole and the emerging standard, and to suggest that this process ignore the vast majority of people is at best naive and at worst extremely damaging to your own position.

Re:HTML5 Video

[Xtifr]

To be fair, the format is entirely open, but patent encumbered.

A bit of an oxymoron there, but I know what you mean. The technical specifications are open; use is not. The latter may not be a factor for the typical home user whose license fee was bundled in with their hardware or OS, but it's going to be a factor for Wikipedia.

Re:HTML5 Video

[Rockoon]

Why? Closed formats don't seem to operate under that constraint. In fact, technical qualities seem to be a non-issue as far as success goes in general.

"Its the money, stupid!"

No, not kickbacks, or payola, or licensing fees.

Lets start at the top. Content providers have been banging their head into the bandwidth wall for a decade, starting when the notion of streaming high quality video really took off. Their cost, primarily, is bandwidth. Their product, primarily, is eyeballs. Their revenue, primarily, is advertisers.

To make this work, they need to offer competitive quality in order to maximize the number of eyeballs, and they need to do it with the least bandwidth in order to offer competitive pricing to advertisers.

H.264 was a big improvement over the previous generation of codecs, which finally allowed what might finally be viable online video streaming businesses.

In this case, technically better still matters... its just about the only thing that matters. These businesses don't have the margin to fuck around. If they drop the ball then they lose their shot at #1.

Re:HTML5 Video

[Korin43]

This would seem like the obvious choice, and it's actually how Safari works (using Quicktime). There are patches for Firefox, but Mozilla refuses to use them because they want to make this a "Theora or nothing" battle, even though they'll never win. Chrome supports Theora and h264, so it hardly matters what their backend is. I think there are patches to webkit to let it use gstreamer as well though. I assume IE9 will use DirectWhatever.
The problem is that Quicktime and DirectShow don't support theora or vorbis by default, so hopefully Mozilla/Wikipedia/anyone else who cares can get them popular enough that Microsoft and Apple have to finally support some free codecs.

Re:HTML5 Video

[shutdown -p now]

No-one said that it is.

The point is that browser remains FOSS. If user is a FOSS purist, he doesn't install the "evil" codecs, and doesn't go to websites which only provide H.264 streams. If user is a pragmatic Linux user, he either buys the codec (freeness of browser not affected), or ignores the law and installs it from "non-US" repositories (legality of browser is not affected).

Most people, of course, just use a mainstream desktop OS, where this all is provided out of the box (and they've paid for it when they purchased their PC/Mac with that OS preinstalled).

Re:HTML5 Video

[qbast]

It's all nice and all, but if open video technology really wants to win, they have to be technically better. There is no other way.

No, actually as long as Theora is not significantly worse than h.264 it does not really matter.

However it's nice to see Open Video Alliance trying to partner with Wikipedia. In addition to being technically better, that's another aspect you need to take care of. You need to make sure websites, TV, phone, computer and so on manufacturers support your technology. You have to work with them to get it supported - not just put it out there and hope it catches up because its "open", because that's not going to happen.

Yeah, just look how popular Vorbis Ogg is in portable music players.

Personally I would also hate to see technically inferior solution being used, as it would eat huge amount of bandwidth. Theora just isn't on the same table with H.264 for Internet video. Theora is based on VP5 from On2 and now that Google acquired them, they're going at VP8.

Actually it is based on VP3 and it is way behind h.264 - it does not even support B-frames! Also being at least one generation behind, Theora is dead end - all that is being done at this point is tweaking the encoder.

What I'm more worried about is that I cannot watch Wikipedia videos with any other device than my PC. Want to see a video clip of a place you're traveling on your phone? Not possible. Want to see videos from Wikipedia with your PS3/360? Not possible.

Well then, make sure you complain to manufacturers of these devices. If enough people care, they will add Theora support.

It will create some serious problems, and I don't think Wikipedia is big enough to push the change alone.

Actually I think Wikipedia is about the only site that can push the change. It is biggest and most popular encyclopedia on the net, has no real competition and would be extremely hard to recreate. There is already bunch of Youtube-like sites, so even if YT switched to some unpopular format, lots of people would just go elsewhere. In case of Wikipedia there is nowhere else to go.

Re:HTML5 Video

[muyshiny]

that's bold of you user #31 million some.

but seriously, it's important and i don't know if you're trolling or what, perhaps i'm that idealistic? please help me understand why having formats and software free of legal entanglement and reducing humanity's dependency on a few, often secretive, organisations is not worth the minimal effort that this volunteer-run collective is taking to contribute to a shared ideology?

we joke and belittle 'slashdot' culture but a lot of this here shit is real sir and i think we do ourselves a disservice. and if you're sincere then damn, go read wikileaks for a bit and see if you can't get a sense of perspective. these are the tools with which we increasingly control our personal identities and the global economy of both ideas and goods. anyone who can understand why it's important to keep that as neutral and transparent as possible really ought to step up wherever they can because... it really is.

Quit embeding the codec support in the browser

[Anonymous Coward]

Let the OS handle it, and let the browser interact via plug-ins.

It's really not that complicated.

Re:Quit embeding the codec support in the browser

[Abcd1234]

Why do you need a package manager when browsers have been facilitating the easy installation of plugins, such as Flash, for years now?

Because the point is the browser shouldn't need *any* codec-specific plugins at all. The browser should simply use the existing mechanisms in the operating system or desktop environment for performing video decoding. On Windows that means DirectShow, and on Linux that means gstreamer. Codec installation is then a task for the operating system or package manager. The result is a better experience for the user, and a simpler implementation for the developers, as they need only to interface to a generic video backend, rather than incorporating a completely codec stack into the browser.

'course, this kind of reasonable design decision would get in the way of pointless political posturing, and who really wants that?

Killer App?

[wjousts]

I'm not sure I see Wikipedia as being the "killer app" for video standards. I'm not sure how many articles would be really enhanced by the addition of video, baring in mind that video would need to be licenced under CC or similar, so clips of TV shows / films would probably be out.

To take a random example (today's featured article) . I'm not sure what video you could usefully add to that article? Especially since somebody who died in 1938 probably isn't featured in many video clips.

And...

[Lunix Nutcase]

And Google, Microsoft and Apple give out a collective *yawn*. Youtube has more traffic than Wikipedia so if Google is pushing H.264 through there it will have far more impact than Wikipedia. Not to mention that Facebook, who also has more traffic than wikipedia and also youtube, also uses H.264 for its video.

the non-free part isn't so bad

[Protonk]

WP should just adopt html5 and give up on the FOSS posturing for once. We already relented on the issue of fair use media--limited use for copyrighted material. Patent protected material seems like a better place to compromise more widely because patents don't live forever. After ~14-21 years, the content path is free. If WP does plan to be around "forever", that isn't too long a time to wait.

Re:the non-free part isn't so bad

[OverlordQ]

They could even host some of the relevant bits of system software and web browser glue-ware.

It does already. Wikipedia uses the OGGHandler extension which tries to determine automagically what method for displaying video the client supports. It supports attempting to use the following clients:

• Cortado (bundled Java applet)
• VLC
• QuickTime with XiphQT
• Totem
• Kaffeine
• KMPlayer
• (ko)GomAudio

And then some more generic support for other cases

Re:Oh, they WILL be paying.

[Lunix Nutcase]

They just won't know it.

Know what? People pay 0 dollars for a browser. Exactly what costs are they bearing due to Apple or Google or Microsoft including H.264 support in the browser?

Oh, and a more idealistic person might even say that they'll not only be paying money

Paying money where? Browsers have all been free for quite some time now.

but paying in a more metaphorical sense with lock-in, etc.

What lock-in? What exactly am I "locked-in" to when I watch H.264 HTML 5 movies on youtube? And how would those movie being encoded in theora make me less "locked-in"?

Wikipedia?

[TopSpin]

So far the comments are focused on teh 3v1Lz H.264 vs. 'open' codecs, why one is better than the other, etc. What about Wikipedia?

Perhaps Wikipedia doesn't actually need to be riddled with video. Maybe Wikipedia is actually better off without it. Have you ever had to suffer through some lengthy, 99% irrelevant video to get a specific piece of information? How many times have you just not bothered to watch that video because it's frustrating, you can't afford the time, don't have just the right version of some plug-in, etc? Ever tried to copy and paste from a video?

How much of the useful content of Wikipedia is going to end up trapped inside videos when easily indexed and searched, entirely unencumbered US Grade-A ASCII^h^h^h^h^hUTF-8 would have been sufficient? How much more bandwidth is Wikipedia going to have to fund to serve up cell phone footage of Silambarasan Rajendar waving at people?

All this means is competition....

[avatar_charlie]

......just not between corporate entities.

No, the competition will be between various wiki-weirdoes over who can be first to enshrine their peckers forever by putting video of it on the articles for "Penis", "Herpes", and any other genital or sex-related article on that site....of which there are no small number.

Re:Oh, they WILL be paying.

[K. S. Kyosuke]

Know what? People pay 0 dollars for a browser. Exactly what costs are they bearing due to Apple or Google or Microsoft including H.264 support in the browser?

I would assume that the licensing fees for MPEG are a part of the Windows and Mac OS X price tag.

Re:This is great for web filtering and searching

[shutdown -p now]

h.264 = Porn and "funny", time wasting, videos

Theora = Actually useful stuff

That's a long way to spell "H.264 wins".