Video Wars, round II
January 13, 2011 11:57 AM   Subscribe

We expect even more rapid innovation in the web media platform in the coming year and are focusing our investments in those technologies that are developed and licensed based on open web principles. To that end, we are changing Chrome’s HTML5 <video> support to make it consistent with the codecs already supported by the open Chromium project. Specifically, we are supporting the WebM (VP8) and Theora video codecs, and will consider adding support for other high-quality open codecs in the future. Though H.264 plays an important role in video, as our goal is to enable open innovation, support for the codec will be removed and our resources directed towards completely open codec technologies. - Google's Chrome is will be joining Firefox in no longer licensing the MPEG-LA H.264 video codec favoured by Apple and Microsoft for use in the HTML5 <video> tag (previously). Not everyone is seeing this as a good thing.
posted by Artw (144 comments total) 8 users marked this as a favorite
 
Huh, everyone here had me convinced that Theora was a total non-starter.
posted by kenko at 12:01 PM on January 13, 2011


Ugh. How long is this going to delay the widespread adoption of HTML 5 video over flash video players?
posted by thsmchnekllsfascists at 12:03 PM on January 13, 2011


Who cares. Every video site will serve both codecs.

This is a non-controversy.
posted by empath at 12:04 PM on January 13, 2011 [1 favorite]


What I want to know is what this means for me as an end-user. I'm not talking about the adoption of the technologies across the web or the cost in effort and tech to upload videos. I'm talking about what this means in terms of how I view videos on Chrome vs. Safari vs. Firefox. If I just have to pick up a plugin on Chrome or something, that's annoying, but I can live with it. If the web is suddenly going to be broken in weird ways because of it when I use Chrome, I may need to rethink my browser choices.

(I expect that what it means to me ultimately is "nothing" and the burden is all on the production end, but that's not clear to me from the initial reactions.)
posted by immlass at 12:06 PM on January 13, 2011


For a little context, Chrome and Firefox hold about 42% of the browser market, whereas IE and Safari hold about 50%. I believe Opera is likely to join in on the WebM side, if it hasn't already, which adds another 2% or so. Basically this makes for a fairly even split.

Personally I view Google's actions here pretty cynically. I think Google's long-term plan here is to switch YouTube over to WebM in order to bring more people over to Chrome. That's the same kind of leveraging technique that people decried when Microsoft used it.

Who cares. Every video site will serve both codecs.

Except only one of the major web browser makers also runs one of the biggest video sites on the internet.
posted by jedicus at 12:08 PM on January 13, 2011 [4 favorites]


Ogg Theora is (unlike Ogg Vorbis, which is technically awesome) - whatever you want to use it for, WebM is almost certainly better quality and lower file size (and will only become more so).

It was never about the video quality though, it's about freedom. WebM ticks the most crucial box.
posted by jaduncan at 12:08 PM on January 13, 2011


"Personally I view Google's actions here pretty cynically. I think Google's long-term plan here is to switch YouTube over to WebM in order to bring more people over to Chrome. That's the same kind of leveraging technique that people decried when Microsoft used it."

Unless, of course, MS just implemented it.
posted by jaduncan at 12:09 PM on January 13, 2011 [1 favorite]


Except only one of the major web browser makers also runs one of the biggest video sites on the internet.

The odds of Google shutting out 50% of the browser market are pretty slim, i'd say. Google has given no indication of being apple.
posted by empath at 12:10 PM on January 13, 2011 [2 favorites]


Overall, it seems like a good idea to me. They're trading away a short-term advantage to try to improve the overall health of the open source ecosystem, which Google is very dependent on.

Ars goes on and on about how 'open' h264 is, but that's only true in small installations and if you don't have the temerity to make any money. There's an awful lot of open source out there that has shipped in excess of 100,000 copies. Pretty much everything in a default Ubuntu install would be one concrete example.

Further, h264 is patent-encumbered, meaning that Apple can potentially change the rules or shut down open source applications that are infringing too closely on things it wants to do.

Google is thinking over the long term; they don't want to enshrine any of kind of standard that prevents people that are using their technologies from making money in any way, and in any amount, they choose. I agree with them that this goal is worth trading away an advantage in the present.

Neither WebM nor Theora are as efficient as h264, but bandwidth expands constantly, and optimizing for good results on today's networks, trading away a thriving ecosystem on tomorrow's, strikes me as typical corporate 1-quarter-at-a-time thinking. We're not building for Q2 2011, we're building for 2020.
posted by Malor at 12:12 PM on January 13, 2011 [11 favorites]


I really don't think this is that much of a big deal. Google seems to be hedging their bets in case MPEG-LA decides to be a Dickasaurus Rex. Certainly Google and Firefox can afford the $6.5mil cap, but what happens when MPEG-LA decides that YouTube, or some future service, violates their licensing contract and demands they stop delivering content with their codec? Or any other bullying that large media companies are notorious for? Google has a big Fuck You in their pocket, and as YouTube is the largest video site on the web, they have considerable pull into making a big Fuck You statement.

Of course this will probably never happen and go the same way as the vague threats cable stations pull every time their contract is up for renewal through Comcast/TWC. 2015 comes up and MPEG-LA wants some regressive DRM built into any client? Go ahead, try it, FireFox and Chrome are already built around a royalty free codec. Do you really want to alienate up to 50% (in Europe at least) of the installed browser base? There's no need for the companies to scramble around and throw in a last minute plugin for Theora when it has been baked in for the last 5 years.

I have no doubt that FireFox and Chrome will make this as transparent to the end user as possible, just as they do with Flash now. I don't think anyone, except perhaps the most paranoid Internet users, think that the two underdogs in the browser game are going to do anything to alienate people.
posted by geoff. at 12:13 PM on January 13, 2011 [5 favorites]


Ugh. How long is this going to delay the widespread adoption of HTML 5 video over flash video players?

That will happen when Ie8 is dead which will happen when XP is dead which is way beyond the timescale where any web dev building anything right now should worry about it.
posted by Artw at 12:14 PM on January 13, 2011 [1 favorite]


Don't worry, this will all be moot in 2028 when the patents expire.
posted by RobotVoodooPower at 12:14 PM on January 13, 2011 [3 favorites]


Unless, of course, MS just implemented it.

Apple and Microsoft are both pretty heavily invested in H.264. They're both MPEG LA licensors, for example. They have economic reasons for wanting H.264 to succeed. Google is basically saying "risk losing browser market share or admit that we can effectively dictate standards and file formats." Whether MS and/or Apple implement WebM or not they stand to lose something.

That said, Google has a bit of an uphill battle, here. H.264 is pretty entrenched on the video production side, so Google's main leverage is on the distribution and consumption sides. That means that at some point a lot of video is going to get transcoded from H.264 to WebM. How lossy is that likely to be?
posted by jedicus at 12:15 PM on January 13, 2011 [1 favorite]


Who cares. Every video site will serve both codecs.

This is a non-controversy.


A nontroversy?
posted by thsmchnekllsfascists at 12:16 PM on January 13, 2011 [3 favorites]


I care because I'm looking at the codecs into which we will be encoding our video in the coming years. Some of the specs are mind-numbing enough as it is, and then finding out how ... underdeveloped the actual server software is kind of freaks me out. I'm still thrashing through documents trying to find out if a single H.264 file can have multiple bitrate streams or something equivalent, with bitrate negotiation on the server end. Even Real Helix Server has had this for ages, so you would think this would be a Pretty Obvious Feature in a server.

I care because going back to all of the source material in the mezzanine level and re-encoding it from there, remapping all links, and so forth, is a significant CPU and human time investment.
posted by adipocere at 12:16 PM on January 13, 2011 [2 favorites]


That means that at some point a lot of video is going to get transcoded from H.264 to WebM. How lossy is that likely to be?

From the recesses of my mind, I remember YouTube at one point talking about how they keep the original file and then encode it into multiple formats as needed. So there's no transcoding at all, at least assuming the original file itself encoded into a lossy format.
posted by geoff. at 12:19 PM on January 13, 2011 [1 favorite]


Moving YouTube to HTML5 WebM would have no effect on the browser market: once Internet Explorer 9 is released, every major browser except Safari will natively support WebM. Granted, that's a big iOS-shaped hole on the mobile end, but it's not like iOS has browser choice anyway. After IE9 WebM will have much wider support (IE, Firefox, Chrome, Opera) than h.264 (IE, Safari).
posted by skymt at 12:19 PM on January 13, 2011 [1 favorite]


Open source meets open sores.
posted by blue_beetle at 12:20 PM on January 13, 2011 [1 favorite]


The odds of Google shutting out 50% of the browser market are pretty slim, i'd say.

Then why bother with WebM? If Google is going to happily present side-by-side H.264 and WebM streams on YouTube, then it's not at all clear to me what their strategy is here. Seems like a significant duplication of effort to no real purpose. It'd be like Flickr serving up PNGs and JPGs side-by-side depending on the browser or something.

Google has given no indication of being apple.

Yes and no. Their methods differ somewhat, but the end goal is the same for both. Apple wants to people to use a vertically integrated set of Apple products and services. Google wants people to use a vertically integrated set of Google products and services.

Further, h264 is patent-encumbered, meaning that Apple can potentially change the rules or shut down open source applications that are infringing too closely on things it wants to do.

Apple cannot do that unilaterally. That's one of the main points of forming a patent pool and using standard licensing terms.
posted by jedicus at 12:20 PM on January 13, 2011 [1 favorite]


Who cares. Every video site will serve both codecs.

I doubt it. The video industry has gotten behind h.264 pretty heavily. It's in BluRay, your phone, your flip video camera, everywhere.

WebM hardware is purely hypothetical at this point and since Chrome will continue to support Flash - and Flash supports h.264, there's no reason to double encode for the roughly 10% of web users on Chrome.

(Adobe stands the most to gain from this.)
posted by device55 at 12:20 PM on January 13, 2011


What I want to know is what this means for me as an end-user. I'm not talking about the adoption of the technologies across the web or the cost in effort and tech to upload videos. I'm talking about what this means in terms of how I view videos on Chrome vs. Safari vs. Firefox.
Practically nothing. Content providers are already encoding most video twice (high-quality H.264 for desktop, low-quality H.264 for mobile); now they'll be encoding it into high-quality WebM (for desktop) and low-quality H.264 (for mobile and IE6). The only difference you'll notice is that watching large YouTube videos won't turn your laptop into a griddle.

nerdy details:

Content served <video> tag is currently divided into Theora (open), WebM (open), and H.264 (patented). Theora is widely supported (90%), but considered somewhat outdated. H.264 is higher-quality, but supported by only 10% of modern browsers. H.264 is backed by Apple and Microsoft, WebM is developed by Google, and Theora is developed by various open-source developers.

Currently, WebM is at 32%, and is only supported in Chrome, Opera, and beta Firefox. When Firefox 4 is released in February, WebM's market share should quickly match Theora's as users upgrade. Eventually, everybody using a modern browser (excluding Safari and IE9) will be able to watch video without Flash.
posted by John Millikin at 12:20 PM on January 13, 2011


From the recesses of my mind, I remember YouTube at one point talking about how they keep the original file and then encode it into multiple formats as needed. So there's no transcoding at all, at least assuming the original file itself encoded into a lossy format.

I'm talking about the case where the original file is H.264-encoded already (i.e. it was recorded using a camera that encodes to that format or edited with software that exports in that format).
posted by jedicus at 12:21 PM on January 13, 2011


Good. If companies had stood up for open standards thirty years ago, computing would look very, very different today... and a hell of a lot better.
posted by vorfeed at 12:22 PM on January 13, 2011 [2 favorites]


h.264 is an open standard. It's just not a free standard.

Google's use of the term 'open' is disingenuous.
posted by device55 at 12:25 PM on January 13, 2011 [3 favorites]


Google wants people to use a vertically integrated set of Google products and services.

I really don't think this is the case. I think they just want enough market share so they can be sure that google won't get shut out by a company like microsoft or apple deciding to start their own search engine.
posted by empath at 12:28 PM on January 13, 2011


I'm talking about the case where the original file is H.264-encoded already (i.e. it was recorded using a camera that encodes to that format or edited with software that exports in that format).

Tough to analyze this is general terms, but there are a bunch of mediocre closed-source H.264 encoders that have inexplicable traction with large commercial software development firms. The file may look sufficiently bad by the time it gets to videosite.com that recompressing with VP8 doesn't do anything noticeable to it.
posted by Inspector.Gadget at 12:30 PM on January 13, 2011


For the purposes they intend the term for -- avoiding encumbrance on all open source video players -- "not open" and "not free" are pretty much interchangeable terms. It's very unsafe to implement h264 in software you intend to give away... as soon as you exceed 100K copies, you're on the hook for license payouts, even if you didn't make any money.

It means that video software can't really be free in all senses of the word -- any popular application would have to monetize its user base to pay off Apple.

I would much rather see a world where free video apps can stay free, no matter how successful they are.

Plus, the terms are only guaranteed until 2015, at which point Apple could really put the screws to free software.

It's a trap, and Google has recognized the trap and is trying to avoid it.
posted by Malor at 12:33 PM on January 13, 2011 [7 favorites]


Yes and no. Their methods differ somewhat, but the end goal is the same for both. Apple wants to people to use a vertically integrated set of Apple products and services. Google wants people to use a vertically integrated set of Google products and services.

I don't think that's quite accurate. Google wants people to use Google products and services. They may structure those products so that they lend themselves to a vertically integrated setup, but they do not demand that to function. Google would rather someone use a 50% Google setup than a 0% Google setup, where Apple seems to only be interested in a 100% Apple clientbase.

I use a lot of Google services, but I do my internet in Firefox. Google does not deny me anything on that basis, and I don't expect that to change. Apple already does, and I expect that to continue.
posted by kafziel at 12:34 PM on January 13, 2011 [2 favorites]


I really don't think this is the case. I think they just want enough market share so they can be sure that google won't get shut out by a company like microsoft or apple deciding to start their own search engine.

Really? Let's see. Google offers a browser, an office suite, email, video sharing, video rentals, image editing and sharing, GPS and maps, ebooks, a mobile operating system, a desktop operating system (in development), instant messaging, VoIP, web search, desktop search, blogging services, online payment services, advertising, an RSS reader, social networking, programming languages, news services, and mobile phone hardware.

"Google’s mission is to organize the world‘s information and make it universally accessible and useful." Its strategy for doing that involves having all of the world's information flow through Google products and services. That means offering a vertically integrated set of products and services, which is basically Apple's strategy.
posted by jedicus at 12:36 PM on January 13, 2011 [2 favorites]


This whole thing is silly.

"Oh no, I'll have to implement multiple versions of video!" Like we don't do that already?

"Oh no, Google is trying to control video with WebM!" As if Apple hasn't been doing the same by refusing to allow Flash on iOS devices?

My favorite was the people who are saying they'll uninstall Chrome in order to use Firefox, which isn't supporting H.264 because of the patent issues.

Last year there were huge questions about whether MPEG-LA would try to actually monetize the H.264 patent and in effect freeze out Firefox. Google offers a perfect solution -- an almost-as-good codec that is open -- and everyone freaks out.

H.264 will just have to be implemented in Chrome (and Firefox) as a plugin, just like Flash. And in a way, that seems fairer. Other than WebM, Google isn't trying to dictate the terms of the web, and even with WebM they're really not. Apple, OTOH, in freezing out Flash IS trying to reshape the web in their own image. And by the way, Apple is doing it with a half-baked and incomplete standard (HTML5) that everyone is treating more like a buzzword than a real advance (esp. given how much CSS3 is getting called HTML5).

Google just told Apple no, you don't get to define the web. We all do. If Apple wants to keep doing it, they'll have to make a less crappy browser than Safari and maybe buy out the MPEG-LA's patent.
posted by dw at 12:36 PM on January 13, 2011 [3 favorites]


It's very unsafe to implement h264 in software you intend to give away... as soon as you exceed 100K copies, you're on the hook for license payouts, even if you didn't make any money.

That depends on the language of the MPEG LA agreement. The summary of the agreement uses the term 'sold.' Free and open source software downloaded for free may not qualify.

It means that video software can't really be free in all senses of the word -- any popular application would have to monetize its user base to pay off Apple...Plus, the terms are only guaranteed until 2015, at which point Apple could really put the screws to free software.

Quit saying 'Apple.' Apple has no power to do any of that unilaterally. It is one member among many of the MPEG LA h.264 pool. It's likely not even the most important member, since most of the MPEG LA revenue comes from stuff like Blu-Ray players, and Apple doesn't care much for Blu-Ray.
posted by jedicus at 12:39 PM on January 13, 2011 [1 favorite]


Google would rather someone use a 50% Google setup than a 0% Google setup, where Apple seems to only be interested in a 100% Apple clientbase.

Apple's approach pushes harder for complete integration, but Google would also rather someone use a 100% Google setup than 50%, and WebM is part of that.

"Oh no, Google is trying to control video with WebM!" As if Apple hasn't been doing the same by refusing to allow Flash on iOS devices?

Apple's opposition to Flash has little to do with video codecs. Flash supports H.264, for starters.
posted by jedicus at 12:44 PM on January 13, 2011


Then why bother with WebM?

If you are a pure consumer, then this decision means nothing to you. You'll continue to just pay money and consume media. If you are a creator of things, though, this means that video now has a broadly adopted standard which you do not have to ask permission or agree to any private entity's onerous terms to use.

That's a big deal.
posted by mhoye at 12:46 PM on January 13, 2011 [2 favorites]


Good on Google for doing this. Open standards are essential for a healthy web. A protocol or codec under patent protection protection isn't open.
posted by formless at 12:47 PM on January 13, 2011


Unfortunately it doesn't change anything in the commercial landscape and it seems more like a minor victory for command-line Linux junkies than anything remotely relevant to the media landscape. HTML tags will still be passed over for other implementations that are cross-browser compatible. Like empath said most programmers will have to go through the hassle of browser detection and proper display anyway. The next-big-deal in terms of physical media will still be decided by Toshiba and Panasonic and Sony and Microsoft.

It's too bad Google's stance is "we *only* support open codecs going into the future"

"Forward-thinking//innovation" does not automatically equal "usable product"
posted by Khazk at 12:47 PM on January 13, 2011


Sun/Oracle should be a big fat warning to folks as well.
posted by Artw at 12:48 PM on January 13, 2011 [3 favorites]


I think everyone is forgetting that WebM is not necessarily safe from patent issues. To quote Jason Garrett-Glaser, the current primary x264 developer:
Finally, the problem of patents appears to be rearing its ugly head again. VP8 is simply way too similar to H.264: a pithy, if slightly inaccurate, description of VP8 would be “H.264 Baseline Profile with a better entropy coder”. Even VC-1 differed more from H.264 than VP8 does, and even VC-1 didn’t manage to escape the clutches of software patents. It’s quite possible that VP8 has no patent issues, but until we get some hard evidence that VP8 is safe, I would be cautious. Since Google is not indemnifying users of VP8 from patent lawsuits, this is even more of a potential problem. Most importantly, Google has not released any justifications for why the various parts of VP8 do not violate patents, as Sun did with their OMS standard: such information would certainly cut down on speculation and make it more clear what their position actually is.

But if luck is on Google’s side and VP8 does pass through the patent gauntlet unscathed, it will undoubtedly be a major upgrade as compared to Theora.
posted by Rhomboid at 12:52 PM on January 13, 2011 [3 favorites]


This feels like remedial school for web browser development. How can the fun of the battles 10 years ago be so easily forgotten?
posted by Fezboy! at 12:54 PM on January 13, 2011 [1 favorite]


I'm talking about the case where the original file is H.264-encoded alread

All videos are transcoded to a standardized set of formats. h264 uploads still get re-encoded to match our formats. So there's no difference.

YouTube will have videos in h264 and WebM for the foreseeable future, as people have surmised (and Flash support isn't going anywhere, of course).

(I can't comment on the meat of this post...)
posted by wildcrdj at 12:59 PM on January 13, 2011


Remember back when Compuserve started enforcing the LZ patent against GIFs? And how half the major browser manufacturers removed GIF support because it was non-free and for awhile we had to use lossy JPGs to encode our text mastheads and it looked like crap? Fortunately PNG finally took over and we used that, except of course PNG support wasn't consistent in all browsers for awhile.
posted by Nelson at 1:01 PM on January 13, 2011


I really don't think this is the case. I think they just want enough market share so they can be sure that google won't get shut out by a company like microsoft or apple deciding to start their own search engine.

"Now who's being naive, Kay?"
posted by entropicamericana at 1:04 PM on January 13, 2011 [1 favorite]


>This whole thing is silly.

>>"Oh no, I'll have to implement multiple versions of video!" Like we don't do that already?

Yeah we do that already, so how does this help us? Oh right, it makes it worse. The browser manufacturers and these huge multi-billion dollar corporations can jockey for position in the world. You know what, I've stopped caring.

It's akin to the Russian empire: Google, Apple, Microsoft, Adobe are the tsars and we are their serfs, forced to waste our time encoding shit because they have no urge to get it together.

And yes the serfs encoded shit in the 19th century, they also killed themselves due to the misery. Which is what will happen to all sane web designers/developers who have to deal with this crap. Either that or revolution! cause that worked out well for the russians, right?
posted by jeremias at 1:07 PM on January 13, 2011


I really wish Apple would release a decent search engine so I could rid myself of Google. They're an advertising company, kids, and you're the product. Just because they wrap themselves in ugly design and say the word "open" a lot doesn't mean they're the good guys.
posted by entropicamericana at 1:08 PM on January 13, 2011 [5 favorites]


Google's use of the term 'open' is disingenuous.

Nuts to that. It is on us as programmers to stop playing these Orwellian language games. Just because a word has been assigned a meaning, it doesn't mean it's right. Consumers hear "open" and think this is a nerd-fight over to completely equivalent things, but it's not even close. Your house is "open": it has windows I can look in. Doesn't mean I can take your TV or hump your dog. Not that I'd do that. I already have a TV.
posted by yerfatma at 1:15 PM on January 13, 2011


Bing is just across the interweb street.
posted by Artw at 1:16 PM on January 13, 2011


I think everyone is forgetting that WebM is not necessarily safe from patent issues.
Neither is H.264; submarine patents apply to every format equally. If you want to avoid any format that might be covered by unknown patents, you'll have to upload videos in .bmp.
Remember back when Compuserve started enforcing the LZ patent against GIFs? And how half the major browser manufacturers removed GIF support because it was non-free and for awhile we had to use lossy JPGs to encode our text mastheads and it looked like crap? Fortunately PNG finally took over and we used that, except of course PNG support wasn't consistent in all browsers for awhile.
Browser vendors never removed .gif support, because Unisys's patents only applied to the compression process.
posted by John Millikin at 1:17 PM on January 13, 2011 [4 favorites]


Bing is just across the interweb street.

Not interested.
posted by entropicamericana at 1:18 PM on January 13, 2011


I really wish Apple would release a decent search engine so I could rid myself of Google.

Wow, thanks for letting us know all about how heartless and cynical Google is. Apple is our one safe port in a storm, huh?
posted by yerfatma at 1:20 PM on January 13, 2011 [2 favorites]


Did I say that?
posted by entropicamericana at 1:21 PM on January 13, 2011


I am glad that Apple is helping to kill Flash. Somebody had to draw a line in the sand or we'd never rid ourselves of the beast. I can't believe how long we've put up with it.

I'm glad that Google and Firefox are helping to kill h264. Somebody had to do it, and it was them. They had to do it. And they did.
posted by Galaxor Nebulon at 1:25 PM on January 13, 2011 [1 favorite]


The most immediate effect is that more sites are going to stick with good ol' proprietary flash, and that's certainly not good for the openness of the web.
posted by gyc at 1:25 PM on January 13, 2011 [3 favorites]


Theora/WebM would have been adopted for HTML5 sooner, but was hindered by price/performance considerations for online/mobile delivery, somewhat different price/performance considerations for hardware due to reliance on (then-expensive) high-precision math for operation of the decoder, and anxiety about submarine patents. A brief flashback here, with many relevant links. The mobile landsscape has changed substantially since then, not least due to the emergence of Android.
posted by anigbrowl at 1:27 PM on January 13, 2011


That's the same kind of leveraging technique that people decried when Microsoft used it.

Well, no, because by using a format that anyone is free to implement encoders and decoders for, Google is explicitly not locking out any competition. For example, Firefox 4 is getting WebM support.
posted by a snickering nuthatch at 1:29 PM on January 13, 2011


The next-big-deal in terms of physical media will still be decided by Toshiba and Panasonic and Sony and Microsoft.

Physical media? How quaint.

The "next-big-deal" is likely to be the last-big-deal.
posted by JaredSeth at 1:30 PM on January 13, 2011


Neither is H.264; submarine patents apply to every format equally.

First, submarine patents are essentially nonexistent these days. There are only about 600 potential submarine applications left, and chances are (since they were filed before June 7, 1995) very few if any apply to (this kind of) video compression.

And patents (submarine or otherwise) do not apply equally. An absolutely enormous amount of due diligence work went into the MPEG LA portfolio, and h.264 is supported directly by a laundry list of licensors that would present an extremely formidable opponent in patent litigation. And even if someone successfully sued over h.264, MPEG LA could simply raise the royalty rates or divert revenue to cover it.

On the other hand, Google is the only major backer of WebM, not counting Microsoft, who would probably be happy to drop it in the event of a patent suit. So while Google's no slouch it is an easier target. And if the suit were successful, either Google would have to eat the whole cost of the resulting royalty or they'd have to start charging for the use of WebM one way or another, which would destroy its major competitive advantage.
posted by jedicus at 1:34 PM on January 13, 2011


I am glad that Apple is helping to kill Flash. Somebody had to draw a line in the sand or we'd never rid ourselves of the beast. I can't believe how long we've put up with it.

It is kind of funny given how long Apple tried to get Quicktime to work in everyone's browser via plugins, and how awful it was on Windows.
posted by smackfu at 1:36 PM on January 13, 2011 [1 favorite]


submarine patents apply to every format equally

But the only reason that VP8 is even being considered is that it is supposedly free of these issues. From every other standpoint -- picture quality, encoding efficiency, encoder/decoder maturity, accelerated hardware support, installed base -- it's inferior to h264. If h264 were not patent-encumbered there would be no debate about the issue whatsoever. If it turns out that some patent troll somewhere discovers that there's some aspect of h264 that VP8 did not sufficiently nerf/remove/rewrite then it loses its only advantage and this whole thing becomes a waste of time. And the only thing we have so far reassuring us that this is not going to happen is Google's lawyers.
posted by Rhomboid at 1:38 PM on January 13, 2011 [2 favorites]


I really wish Apple would release a decent search engine so I could rid myself of Google. They're an advertising company, kids, and you're the product.

Google is a publishing company. They are no more an advertising company than a magazine publisher is; you are confusing their business model with their business. An advertising company creates advertising; Google does not, but rather sells ad placement services. Little-to-no information about Google's user base is handed back to the advertiser; if you doubt this, set up an adwords account (I think they give you a few $ of free credit to experiment with) and see how little information you really get about who's viewing it. Now compare that to the amount of target specification you can deploy on some competing services.
posted by anigbrowl at 1:40 PM on January 13, 2011 [6 favorites]


It is kind of funny given how long Apple tried to get Quicktime to work in everyone's browser via plugins, and how awful it was on Windows.

Quicktime was awful period. Flash video killing Quicktime as a major web video format was a godsend.
posted by kmz at 1:41 PM on January 13, 2011 [4 favorites]


Well, no, because by using a format that anyone is free to implement encoders and decoders for, Google is explicitly not locking out any competition. For example, Firefox 4 is getting WebM support.

It's free now, but there's nothing legally binding Google to keep it free forever. I think that's what people are worried about -- it'll get huge adoption and then Google will crank up the pain. The first hit is always free, etc etc.
posted by breath at 1:45 PM on January 13, 2011


I hate it when mommy and daddy fight!

WebM can be the greatest thing since sliced bacon but it doesn't matter if there isn't device and OS support. I'm doubting WebM will be viewable on my iPhone anytime soon. (I know that's because Apple are a bunch of dicks and don't want a Chrome/Firefox alternative on iOS devices...but I want to play it on Safari on my iPhone!).

Compared to flv or wmv files, on my Mac H.264 is a dream come true. It "just works" without the spinning beachball making an appearance first.

Drawing this line in the sand now, appears to be a little premature, unless next week, Google's YouTube division will have completed all of its shit to WebM. So Chrome/ChromeOS devices won't get support for H.264 even though YouTube spent zillions supporting the format. Will Hulu and Netflix fall over themselves doing that work.

It isn't like the Chrome team is two guys in a garage, I would say they could afford to supply engineering talent to WebM and H.264 at the same time. When there's actual content and the it is making headway, then pare it back. Then drop it eventually.

I was surprised by this because in the past Google was all about doing everything to provide eyeballs for the ads its served, or getting us to be within the Googleverse to see more ads. This is why it makes perfect sense to have support for iOS or Blackberry, Palm, in Maps, Goggles, Voice, etc. Yes, Google may compete with the other smartphone OSes but they want to make sure they're the go-to app.

Considering Google doesn't make a dime on licensing Chrome, Chrome OS or Android. And even the "Googlephones" are made by others. The whole project is really just another way to get people back to google and generate ad revenue.

It just seems a very old fashioned approach I'd expect from Microsoft or Adobe. One thing for certain I think is Flash video will go the way of the Real Player.
posted by birdherder at 1:45 PM on January 13, 2011 [1 favorite]


And for the record my take on this is that the real villains in this story are the bureaucrats and/or politicians that decided to allow software algorithms to be patentable. If we could somehow right this vile injustice and live in a sane world then none of this would matter, and we'd be able to freely chose a standard based on criteria that actually matter instead of this licensing nonsense.
posted by Rhomboid at 1:46 PM on January 13, 2011 [7 favorites]


I never had a problem with Quicktime in those dark days that I was a Windows user. (95-XP) *shrugs*
posted by entropicamericana at 1:46 PM on January 13, 2011 [1 favorite]


Compared to flv or wmv files, on my Mac H.264 is a dream come true.

Flv (and wmv for that matter) is a container format, not a video format. Youtube serves you h264-in-flv, at least if you use the Flash version of their site.
posted by Rhomboid at 1:49 PM on January 13, 2011 [4 favorites]


Chrome/ChromeOS devices won't get support for H.264 even though YouTube spent zillions supporting the format

Well, they won't get h264 support through HTML5. Chrome and ChromeOS both support (in fact, come with built in) Flash. Which supports h264.
posted by wildcrdj at 1:57 PM on January 13, 2011


The difference between WebM and H.264 isn't that the second is "patent encumbered" and the first isn't: it's rather that we know (more or less) which patents are involved in H.264, and who to pay for the licenses under which conditions, whereas we don't have the slightest idea whatsoever of which patents may be infringed by WebM.
posted by Skeptic at 2:08 PM on January 13, 2011 [1 favorite]


I feel liked I just stumbled into a gladiator's arena on a distant planet, where dinosaurs duke it out in front of howling alien spectators, and I'm tiny and trying not to get crushed underfoot. I have learnt much here.
posted by stonepharisee at 2:25 PM on January 13, 2011 [1 favorite]


And patents (submarine or otherwise) do not apply equally. An absolutely enormous amount of due diligence work went into the MPEG LA portfolio, and h.264 is supported directly by a laundry list of licensors that would present an extremely formidable opponent in patent litigation. And even if someone successfully sued over h.264, MPEG LA could simply raise the royalty rates or divert revenue to cover it.
MPEG-LA does not provide indemnity against patent suites. If I license H.264 from them, and it turns out that H.264 infringes on some third party's patents, that third party will be able to sue me for damages without MPEG-LA lifting a finger.

In other words, from a patent perspective, WebM is strictly better than H.264. Both might be covered by unknown patents, but only H.264 is also covered by known patents.
It's free now, but there's nothing legally binding Google to keep it free forever. I think that's what people are worried about -- it'll get huge adoption and then Google will crank up the pain. The first hit is always free, etc etc.
Google has granted a perpetual, royalty-free patent license to VP8 (WebM's codec). There's no take-backsies.
posted by John Millikin at 2:42 PM on January 13, 2011 [4 favorites]


MPEG-LA does not provide indemnity against patent suits. If I license H.264 from them, and it turns out that H.264 infringes on some third party's patents, that third party will be able to sue me for damages without MPEG-LA lifting a finger.

1. Google doesn't indemnify WebM users, either.

2. Chances are the MPEG LA members and major licensees have much, much deeper pockets than you do and have accrued far more in the way of damages. Thus, it's more likely that a hypothetical patentee would sue one of them rather than some startup or open source project.

3. It's in MPEG LA's interest to ensure that everyone with applicable patents is in the patent pool and that anyone who won't play ball gets their patents invalidated. That's one of the main reasons for having a pool in the first place. So even if this hypothetical patentee started suing only small players in the H.264 world, MPEG LA might still intervene in a number of ways.

In other words, from a patent perspective, WebM is strictly better than H.264. Both might be covered by unknown patents, but only H.264 is also covered by known patents.

It's likely that far more due diligence went into ensuring that every relevant patent is in the H.264 pool than went into clearing WebM.
posted by jedicus at 2:53 PM on January 13, 2011


1. Google doesn't indemnify WebM users, either.
I never claimed they did. My point is that if you're worried about being sued for patent infringement, MPEG-LA's portfolio is irrelevant -- the safety of VP8 and H.264 against unknown patents is equal.
2. Chances are the MPEG LA members and major licensees have much, much deeper pockets than you do and have accrued far more in the way of damages. Thus, it's more likely that a hypothetical patentee would sue one of them rather than some startup or open source project.
Exactly why any patent owner will come after me, and other small fry -- I don't have the money to defend myself. They can build up a few wins before going after the money. That's why patent trolls often start off by suing small companies, so they can get precedent on their side.
3. It's in MPEG LA's interest to ensure that everyone with applicable patents is in the patent pool and that anyone who won't play ball gets their patents invalidated. That's one of the main reasons for having a pool in the first place. So even if this hypothetical patentee started suing only small players in the H.264 world, MPEG LA might still intervene in a number of ways.
In other words, you're relying on MPEG-LA's goodwill to bail you out of potential patent lawsuits?
It's likely that far more due diligence went into ensuring that every relevant patent is in the H.264 pool than went into clearing WebM.
According to what evidence?

The Xiph foundation has performed extensive due diligence to ensure Theora (VP8's predecessor) is free of patent threats. Google has paid over a hundred million USD to acquire On2, and altered billion-dollar product lines on the belief that VP8 is safe -- these actions must have involved substantial due diligence.

There is no reason to believe that Theora or VP8 is covered by any patent; the only entities claiming they are belong to MPEG-LA itself.
posted by John Millikin at 3:08 PM on January 13, 2011 [6 favorites]


smackfu: "It is kind of funny given how long Apple tried to get Quicktime to work in everyone's browser via plugins, and how awful it was on Windows"

I just hit a web page recently that wanted me to install Quicktime. I hadn't thought about that horror in ages. In fairness, it wasn't nearly as bad as RealAudio but still pretty sucky. As much as people harp on Flash, it still works 10x better than what came before it.
posted by octothorpe at 3:41 PM on January 13, 2011


Exactly why any patent owner will come after me, and other small fry -- I don't have the money to defend myself. They can build up a few wins before going after the money. That's why patent trolls often start off by suing small companies, so they can get precedent on their side.

That's so misinformed it's not even wrong. If a patentee sues a small company because it can't defend itself, then that small company is just going to cave and settle. Heck, over 95% of patent cases settle anyway. Settlements do not build precedent, and a large company is not going to be snowed by seeing a list of tiny companies that settled.

I never claimed they did. My point is that if you're worried about being sued for patent infringement, MPEG-LA's portfolio is irrelevant -- the safety of VP8 and H.264 against unknown patents is equal.

No, they aren't. The H.264 patent pool likely drew off anyone with a relevant patent. Litigation is expensive. Adding your patent to the pool is comparatively cheap and easy. Furthermore, the basic H.264 specification was finalized back in March of 2003. The statute of limitations on patent infringement in the US is 6 years. Anyone with a relevant patent is already losing potential revenue to the statute of limitations and is starting to run up pretty hard against the doctrine of laches.

By contrast, WebM is relatively new. It's based on VP8, which only dates back to September of 2008.

There is no reason to believe that Theora or VP8 is covered by any patent; the only entities claiming they are belong to MPEG-LA itself.

That's a contradiction. "There's no reason to believe it's covered by a patent, except for the claims made by people with patents that could very well cover it."
posted by jedicus at 3:46 PM on January 13, 2011


I really wish Apple would release a decent search engine so I could rid myself of Google. They're an advertising company, kids, and you're the product.

Gee, I wonder what Apple could possibly do to make a search engine profitable.
posted by me & my monkey at 4:01 PM on January 13, 2011 [1 favorite]


RealAudio

The horror...
posted by Artw at 4:03 PM on January 13, 2011


That's so misinformed it's not even wrong. If a patentee sues a small company because it can't defend itself, then that small company is just going to cave and settle. Heck, over 95% of patent cases settle anyway. Settlements do not build precedent, and a large company is not going to be snowed by seeing a list of tiny companies that settled.
I never said anything about settlements. If a third party sues me for patent infringement, and I fight, and they win, then they will have precedent to use against larger companies.
No, they aren't. The H.264 patent pool likely drew off anyone with a relevant patent. Litigation is expensive. Adding your patent to the pool is comparatively cheap and easy.
Litigation is only expensive if litigating against somebody with the means to fight back. Look at the RIAA for a counterexample: they are making money by filing many thousands of lawsuits against individuals without the means to mount an equal defense.
Furthermore, the basic H.264 specification was finalized back in March of 2003. The statute of limitations on patent infringement in the US is 6 years. Anyone with a relevant patent is already losing potential revenue to the statute of limitations and is starting to run up pretty hard against the doctrine of laches.
There is no statute of limitations on patent infringement in the US; valid patents are a source of risk until they expire. Laches only applies if the delay is unreasonable -- if some company somewhere has a patent covering H.264 or VP8 , but doesn't realize it for a decade, they can still win suits when they begin enforcement.
By contrast, WebM is relatively new. It's based on VP8, which only dates back to September of 2008.
WebM isn't "based on VP8"; VP8 is a component of WebM. VP8 is the codec, WebM is the codec plus a container based on Matroska.

VP8 itself is derived from older On2 codecs (VP7 and VP6), with a core design which dates back to before H.264 was released. It is similar to Theora, which has been under public development since 2005 and has been extensively researched.
That's a contradiction. "There's no reason to believe it's covered by a patent, except for the claims made by people with patents that could very well cover it."
It's not a contradiction because I never said their claims are true. MPEG-LA have a financial interest in preventing open codecs such as Theora and VP8 from gaining traction, and therefore, in making false claims that such codecs are covered by their patents.

No third-party patent research has ever uncovered evidence of such mythical Theora/VP8 -related patents. It's pure propaganda, along the same lines as Microsoft claiming Linux infringes their patents.
posted by John Millikin at 4:04 PM on January 13, 2011 [7 favorites]


This only applies to the US, effectively - europe doesn't have software patents, nor does most of the rest of the world.

Opera and firefox 4 only support webM because of the patents on h264. With chrome added to that, that adds up to a very substantial market share. Once IE9 comes out, along with firefox 4 - both due in the next few months - that means well over 60% of the desktop browser space will support WebM. The only modern desktop browser that won't be joining in will be safari.

As a video hosting site, you'd have to be *nuts* not to consider webM as an option for HTML5 streaming.

Now, if chrome still supported both, and IE9 supported both, and you're only throwing firefox and opera under the bus by going h264 only, that'd be quite tempting to so - after all, you could just support them with flash, right?

Now, I think we're going to see both webM and h264 support, with flash (h264 embedded) for legacy browsers. Hell, even XP users have a choice of 3 browsers to get webM capability. The pressure is really going to be on apple, as the only main browser not supporting webM, to add support for it. It is free, after all. Given it's apple though, I doubt they'll fold.

So we'll probably see a three way split - webM for everyone but apple, h264 for iOS and maybe safari, and flash for old browsers. It may well mean that some smaller sites decide to just ignore iOS altogether, and support flash and webM. Mobile streaming is still relatively rare, iOS still has a tiny market share of the overall browser market, and with data caps and network providers heading rapidly in the direction of clamping down mobile streaming it's not quite such a non-starter.

The mobile space is of course a much more open situation; hardware accelerated h264 support is still thin on the ground (mostly in high end devices), loads of devices don't support accelerated flash, and of course iOS doesn't support flash at all.

I'm told that it's not that much of a stretch to use the same hardware to accelerate both webM and h264. I think we'll see a lot of devices, especially android, coming out that accelerate both. Except apple devices, of course.

I think the end result is that now apple, and iOS especially, are the holdouts to what is otherwise a common standard, and at risk of marginalisation in the future. XP users will also be under greater pressure to switch away from IE, if they want a modern browser. If google hadn't dropped h264 in chrome - a browser with explosive growth, mostly canibalizing IE, not firefox - instead it would be firefox and opera at serious risk of being pushed to the margins.

The latter would definitely suck for the non-zero percentage of linux desktop users, as flash really sucks there. In terms of market share, linux users are pretty similar to OSX penetration. If chromeOS takes off in the tablet/netbook space*, that will bump up linux/webM only usage dramastically, which again swings the momentum away from apple and towards open standards.

It's a bold move. In the short term it's going to prolong flash as a dominant force, but in the longer term it should help stop open source browsers from being cut off of the future of the web. And stick a fork in apple's eye, which I'm sure Google aren't shedding any tears over.

Purely personally, as an android and linux user, I'm absolutely ecstatic about this. It does remind me of a time that .wmv was marching to take over video streaming, and .wma radio was completely inaccessible to non windows users. h264 was threatening to do the same, and I'm hopeful that we'll see a plural, inclusive future no matter what platform I choose.

* not that I expect chromeOS to get far, but I didn't predict the explosive growth of chrome, nor the mass market appeal of android to people who aren't hardcore techies like me, so we'll see what happens.

And yes, I like google products, and I use them. They give me a lot of options I simply don't get with apple. And they're free of charge. If I want, I can use something else, and still have access to all my own data. The same is not true of microsoft or apple. I can live with the non-monentary cost in order to get that openess.




posted by ArkhanJG at 4:08 PM on January 13, 2011 [3 favorites]


If Google is the only one using WebM, and if people will still serve up H.264 content, the way to view H.264 through Chrome would be through its built-in Adobe Flash Player plug-in.

The irony of this decision is that it helps prop up a closed, proprietary plug-in and format, under the guise of promoting "open"-ness.

I suspect one of the unspoken goals of this decision is to help preserve Adobe Flash and make it a selling point for Android, which ultimately helps push the Android platform.
posted by Blazecock Pileon at 5:20 PM on January 13, 2011 [1 favorite]


I never said anything about settlements. If a third party sues me for patent infringement, and I fight, and they win, then they will have precedent to use against larger companies.

You talked about going after small fry. Those companies aren't going to defend themselves. because it's far too expensive. In the patent context, litigation to a judgment on the merits (i.e., something that could be called a precedent) costs in the hundreds of thousands and up (way, way up). If you really want a proper precedent you need an appellate decision from the Federal Circuit. That will likely mean spending a million or more.

Look at the RIAA for a counterexample: they are making money by filing many thousands of lawsuits against individuals without the means to mount an equal defense.

Actually, they probably aren't. The average settlement is ~$3,000. That buys about 4-5 hours of time from a litigator at a large firm. I would be shocked if they could operate efficiently enough to break even. And that's not counting the hundreds of thousands to millions of dollars spent on the handful of cases that went to trial and appeal.

There is no statute of limitations on patent infringement in the US; valid patents are a source of risk until they expire.

Please stop talking about something you apparently know very little about. "Except as otherwise provided by law, no recovery shall be had for any infringement committed more than six years prior to the filing of the complaint or counterclaim for infringement in the action." 35 USC 286. If you want to get ridiculously technical, this is a limitation on damages rather than a true statute of limitations, but it's often referred to that way.

Laches only applies if the delay is unreasonable

Unreasonable delay is presumed after 6 years. AC Aukerman Co. v. RL Chaides Const. Co., 960 F. 2d 1020 (Fed.Cir. 2002).

if some company somewhere has a patent covering H.264 or VP8 , but doesn't realize it for a decade, they can still win suits when they begin enforcement.

Not if they should have known about the infringement. Actual knowledge is not required. Given how well-known H.264 is, it's likely that any patentee would have a hard time showing that they should not have reasonably known about the infringement.

WebM isn't "based on VP8"; VP8 is a component of WebM.

Ah, in much the same way that Ubuntu is not based on Linux, since the kernel is a mere component of the operating system as a whole. Nonsense.

VP8 itself is derived from older On2 codecs (VP7 and VP6), with a core design which dates back to before H.264 was released. It is similar to Theora, which has been under public development since 2005 and has been extensively researched.

So perhaps some of the older aspects of the codec might be protected by laches or the 6 year limit on damages, but the newer aspects of VP8 (and thus WebM) are vulnerable.
posted by jedicus at 5:40 PM on January 13, 2011


This only applies to the US, effectively - europe doesn't have software patents

Check the AVC/H.264 patent list [pdf]. There are many, many non-US patents on that list, including patents from Europe, Japan, South Korea, China, Mexico, Australia, Taiwan, Canada, Singapore, Hong Kong (which is separate from China, patent-wise), and elsewhere.
posted by jedicus at 5:43 PM on January 13, 2011


Quick, to the phones! Somebody must tell Google that internet lawyer jedicus has discovered a critical and unforeseen flaw in their plan! I curse your name John Millikin, you who have abused poor Google's trust and led them astray with your twisted and warped view of patent law! Oh, if only Google had not, in what can only be called madness, forsworn all legal council except for that of John Millikin!
posted by Pyry at 6:51 PM on January 13, 2011


Perhaps they could just hire him instead.
posted by anigbrowl at 7:00 PM on January 13, 2011


When does a patent infringement occur?

Honest question.

When you develop an infringing product? sell it? distribute it?

Cause my totally non lawyer reading would suggest that the establishment time if the MPEG-LA group would be completely beside the point, since you would re-infringe every time you distributed infringing code.
posted by Greald at 7:35 PM on January 13, 2011


When does a patent infringement occur? When you develop an infringing product? sell it? distribute it?

Development and selling it definitely count: "Except as otherwise provided in this title, whoever without authority makes, uses, offers to sell, or sells any patented invention, within the United States or imports into the United States any patented invention during the term of the patent therefor, infringes the patent." 35 USC 271(a).

I don't know if the courts have decided whether or not distributing an infringing program for free counts as a sale, but so long as the software was actually used (i.e. end users committed direct infringement through use), then the distributor would likely be indirectly liable under 271(b): "Whoever actively induces infringement of a patent shall be liable as an infringer."

Cause my totally non lawyer reading would suggest that the establishment time if the MPEG-LA group would be completely beside the point, since you would re-infringe every time you distributed infringing code.

Yes and no. Yes in that recovery for infringement within the last 6 years would not be barred by the statute. No in that laches could still apply, and no in that recovery would be barred for acts of infringement that occurred more than 6 years ago. Since H.264 has been around for well over 6 years now, every day that a hypothetical plaintiff doesn't sue is money wasted.
posted by jedicus at 8:29 PM on January 13, 2011


You would think Google would be glad to be rid of Flash as that would mean more content they can index and serve ads next to. All they had to do was nothing, and yet they did this instead.
posted by ridogi at 8:35 PM on January 13, 2011 [1 favorite]


If Google is the only one using WebM, and if people will still serve up H.264 content, the way to view H.264 through Chrome would be through its built-in Adobe Flash Player plug-in.

But Google will not be the only one using WebM - Firefox will also use it.

I suspect one of the unspoken goals of this decision is to help preserve Adobe Flash and make it a selling point for Android, which ultimately helps push the Android platform.

That's kind of silly. If Google wanted to preserve Adobe Flash, they could (a) use it on more of their web properties, and (b) not port YouTube to HTML 5.
posted by me & my monkey at 9:04 PM on January 13, 2011


But Google will not be the only one using WebM - Firefox will also use it.

Google has YouTube. Firefox has nothing to use WebM with, in comparison.
posted by Blazecock Pileon at 9:29 PM on January 13, 2011


As long as it's not possible to legally create an open source H.264 enabled browser, H.264 isn't open enough in the important ways. You do not have a open web -- in the free/libre sense -- when its clients can't be freely implemented and re-implemented. Imagine a world where HTML itself was controlled by a patent association that charged fees to anyone who implemented authoring or rendering software and you start to get the idea. Yet somehow certain vendors have apparently gotten away with playing exactly that game with a key piece of the HTML5 spec.

If the members of MPEG-LA want H.264 to be a standard compatible with the core values of an open web, then there's one path and only one path: freely available licensing. Like Google just did.

If they want to protect and mine the value from their intellectual property, that's certainly their privilege, but they can't have their cake and eat it too.

(And regarding Flash... sure, it has its problems, but as far as I can tell, anybody who wants to can implement their own runtime or a compiler targeting Adobe's runtime without a fee or legal challenge.)
posted by weston at 9:51 PM on January 13, 2011


"As much as people harp on Flash, it still works 10x better than what came before it."

I'm guessing you were running Windows? Because on a Mac QuickTime is and has been just fine for a very long time.

Flash on the other hands burns CPU (and therefore battery life) like a bitch. When a simple Flash game causes the fans on 2.8GHz computer to kick in something is wrong.

I wouldn't much care about the H264 vs WebM fight except that it helps bring Flash back from its impending doom. And that sucks. Flash needs to die.
posted by schwa at 10:08 PM on January 13, 2011


(And regarding Flash... sure, it has its problems, but as far as I can tell, anybody who wants to can implement their own runtime or a compiler targeting Adobe's runtime without a fee or legal challenge.)

Like Google, Adobe uses the term "open" when doing so suits its marketing purposes. Occasionally, when "open" conflicts with closed, closed wins, e.g. "Adobe has issued a DMCA removal request for rtmpdump". There was a decent discussion of why Adobe Flash is not open source on Reddit about a year ago, which is worth reading.
posted by Blazecock Pileon at 10:13 PM on January 13, 2011 [1 favorite]


Google has YouTube. Firefox has nothing to use WebM with, in comparison.

I see what you're getting at, now. But publishers are going to follow the supported formats. If everyone can view WebM video, publishers will publish WebM video. And if Google wanted to help Flash, as you proposed earlier, they'd just dispense with all this support for HTML 5 video functionality and leave things the way they were.

Like Google, Adobe uses the term "open" when doing so suits its marketing purposes.

Adobe doesn't really use the term "open" that much in relation to Flash, so I'm not sure what you're getting at here. But the fact is, there are third-party SWF generation tools, and there are free SWF generation tools from Adobe. The SDKs for Flex and AIR are freely downloadable. There are no deployment costs specific to the SWF format.

"Adobe has issued a DMCA removal request for rtmpdump".

RTMP != Flash Player.
posted by me & my monkey at 10:36 PM on January 13, 2011


RTMP != Flash Player.

No, RTMP is not Flash Player, but the point was about the use of language, and a link to discussion follows that focuses specifically on Flash Player.
posted by Blazecock Pileon at 10:50 PM on January 13, 2011


it helps bring Flash back from its impending doom. And that sucks. Flash needs to die.

It does a decent enough job at a lot of what it does, and it still does a number of things beyond simple progressive video playback that the web standards focused technology either does poorly or doesn't do at all. So it'll survive for a while just fine no matter how this fight goes. Though of course this would help 'em a bit.

Of course, it won't be Google's fault really, because that only happens if for some reason Apple and Microsoft don't pick up WebM.

Since there are BSD-licensed libraries out there and Google's issuing a royalty-free license, the investment required to do that would be negligible. I'd also hazard my non-lawyerly guess as members of the H.264 patent pool, they're probably also fairly safe from whatever of the speculated legal risks actually turn out to have any substance at all (though of course I don't know what contract terms they might have agreed to).

So, why wouldn't they add it? All I'm left with is a desire to protect control or profits. Again, their privilege, but if they have to cede some ground back to Adobe as they do it. Fine with me. At this point in the game, I'm more comfortable with that than I am with any more leverage going to Cupertino or Redmond, and the idea that any part of a web standard should fall under licensing requirements is a sin that they should already be suffering for.

Like Google, Adobe uses the term "open" when doing so suits its marketing purposes. Occasionally, when "open" conflicts with closed, closed wins, e.g. "Adobe has issued a DMCA removal request for rtmpdump". There was a decent discussion of why Adobe Flash is not open source on Reddit about a year ago, which is worth reading.

I'm familiar with the issue among others. Like I said, Flash has its problems, including a few of Adobe's simultaneous claims of "openness" surrounding Flash combined with its possessiveness of RTMP and other technologies. But the fact remains that legally unmolested (not to mention actually blessed & official) open source compilers and runtimes for Flash exist, and that's explicitly impossible for H.264 right now, and that's unacceptable for a web standard regardless of how you want to define free/open. And at least Adobe hasn't yet had the gall to claim RTMP is a web standard.

Also whatever other beefs you may have with Google, in the context of this particular conversation, I don't understand how you can imply that they're playing games with the term open. Documentation, BSD licensed code, and a no-takebacks license grant is about all even the most libre-concerned purist could really ask for short of hiring an private army to enact a coup and abolish software patents.
posted by weston at 11:00 PM on January 13, 2011 [1 favorite]


Check the AVC/H.264 patent list [pdf]. There are many, many non-US patents on that list, including patents from Europe, Japan, South Korea, China, Mexico, Australia, Taiwan, Canada, Singapore, Hong Kong (which is separate from China, patent-wise), and elsewhere.

Oh, I'm not disputing that
a) there are software patents granted at the national level in european or other countries, either because they were granted on something that is patentable (i.e. hardware), or the relevant patent office cocked it up.

b) licencees stuck in pretty much anything vaguely related to h264.

However, it has been tested repeated in european courts that the EPO standard applies; they are not enforceable. France banned software patents in 1968, for example.

Any patent action taken by MPEG-LA against webM users will not take place in europe. Well, unless they want to play patent troll, and just try and outspend companies with a case that is unwinnable. European courts are a lot less tolerant of that than US courts though.

Canda:

"1. Unapplied mathematical formulae are considered equivalent to "mere scientific principles or abstract theorems" which are not patentable under section 27(8).

2. The presence of a programmed general purpose computer or a program for such computer does not lend patentability to, nor subtract patentability from, an apparatus or process. "

Japan:

"In Japan, the application of a scientific principle is a prerequisite to patentability. Statutory subject matter includes as inventions "any highly advanced creation of technical ideas by which a physical law of nature is utilized"(Japanese Patent Act, Article 2). If this "cause and effect" relationship is linked by a law of nature, then the subject matter is statutory. If the relationship is a mathematical formula only, then the subject matter is not statutory. Thus, inventions based on human mental activities are non-statutory."

Mathematics alone - i.e. codecs - are not patentable.

Mexico does not specifically allow or deny software patents as far as I'm aware. In Australia, they are partially patentable - mathematics alone is not, but computer implemented (i.e. a hardware/software combo) is. So the codec isn't patentable, but a hardware decoder would be.

South Korea does allow software patents, as does China, as does quite a bit of SE asia.

I stand by my assertion that this is primarily a US problem. I do not see end-user lawsuits in China being a likely outcome.

Of course, this is all based on the idea that webM actually infringes somebodies' patents. Until MPEG-LA actually come up with some examples, I'm currently treating their assertion as pure FUD, just as microsofts claims against linux are FUD.
posted by ArkhanJG at 1:06 AM on January 14, 2011


Of course, it won't be Google's fault really, because that only happens if for some reason Apple and Microsoft don't pick up WebM.

IE9 has webM support and h264, but of course that's vista/win 7 only. The only major browser that doesn't is safari in OSX and on iOS, and IE8 or older. Of course, IE8 and older don't support h264 either.

So if you're an XP user, and want HTML5 video streaming, your choice is firefox, opera or chrome for webM support. safari on windows does have html5 h264 support, but it doesn't work without quicktime installed. Safari on windows is a tiny install base, and even smaller use base. IE8 supports neither.

Given XP still has what, a 50% user install base? That's quite a hefty bunch of people currently using flash that could be tempted with quicker, better video streaming without plugins onto an alternative browser - and your choice is basically webM, webM, or webM.

Windows 7 passed the OSX install base within a month of launch. It's already at 20+%. Vista marketshare is declining (thank god) at around 15%. With IE9, chrome, firefox and opera on those, plus firefox as the defacto default on linux, that easily puts webM available for 90% of the desktop. 100% if you allow for firefox on OSX. (Yes, I'm ignoring the mobile space for the moment).

Alternatively, h264 support is in IE9, and safari. That's 45% desktop space support absolute maximum, assuming every single user of vista/win 7 uses IE9, and every OSX user safari. The only way for h264 support to be bigger than that is people using safari on XP with quicktime; as opposed to their current browser, using flash.

Those are maximums. Now if we add the browser share to that:

45% for firefox, chrome and opera. IE8 and older have say, 45%. Safari makes up 5% at best.

So that's 45% of XP users, 100% of vista/7 users (autoupdates for the win), and say 20% of OSX users with firefox, that's just under a 60% user base of available webM capable browsers.

For h264, we have 50% of win 7/vista users (IE9+safari), 5% of XP (safari) users, and 80% of OSX users. That's just under 33% of the install base.

So with webM, you can reach 60% of the desktop space. 100% are possible if they use firefox, chrome, or opera on OSX or XP.

With h264 you can reach 33% of the desktop space. More is only possible if you convert XP users to safari, or vista/7 users back to IE9 from their current alternative.

As a developer/webhost, webM HTML5 looks awfully tempting as the flash successor.
posted by ArkhanJG at 1:39 AM on January 14, 2011


Google should release a high-performance, synthesizable WebM decoder. Then show the chipmakers how to graft it on to a standard ARM core. That would do wonders for mobile support.

Without a free WebM implementation, I don't think chipmakers are going to include it. Hardware video decoders take a ton of engineering effort. Google doesn't have the buying power to just demand hardware support. They would need to develop the IP and hand it to the chipmakers on a silver platter.
posted by ryanrs at 2:47 AM on January 14, 2011 [1 favorite]


I just hit a web page recently that wanted me to install Quicktime. I hadn't thought about that horror in ages.

m-w.com uses Quicktime for its pronunciation examples. I had to try 3 computers to find one with the Quicktime plugin installed just so I could prove that "gingham" has a hard-g at the start.
posted by smackfu at 5:27 AM on January 14, 2011 [1 favorite]


Android does have webM/VP8 support in 2.3 (gingerbread), and hardware support is included in the Nexus S. ARM, nvidia and broadcom announced plans for webM/VP8 hardware accelerated playback in May last year. ARM at least now have it in production. Qualcom, TI are onboard, and intel will support hardware decoding in atom if it proves popular.

I think high-end handsets in 2011, with the exception of apple, will have webM hardware support as a pretty common feature.
posted by ArkhanJG at 5:33 AM on January 14, 2011


I'm guessing you were running Windows? Because on a Mac QuickTime is and has been just fine for a very long time.

Well fine but ~95% of us don't use Macs.
posted by octothorpe at 6:24 AM on January 14, 2011


Until MPEG-LA actually come up with some examples, I'm currently treating their assertion as pure FUD, just as microsofts claims against linux are FUD.

Neither MPEG-LA nor Microsoft wants to spell out which patents it believes WebM or Linux (respectively) may be infringing because doing so would likely create declaratory judgment jurisdiction. Things didn't used to be so touchy, but the decision in Medimmune v. Genentech made it difficult to even politely suggest the possibility that a given patent might be infringed upon without creating DJ jurisdiction. Generally speaking, patentees do not want to do that, so MPEG-LA and Microsoft keep mum until it's strategically useful to sue.
posted by jedicus at 6:38 AM on January 14, 2011


Neither MPEG-LA nor Microsoft wants to spell out which patents it believes WebM or Linux (respectively) may be infringing because doing so would likely create declaratory judgment jurisdiction

Also, they don't want to say what patents might be infringed because for WebM/Linux you might/absolutely would be able to code around them.
posted by a snickering nuthatch at 8:55 AM on January 14, 2011


Plus, you get guys like jedicus arguing your case for free.... "use the codec you have to pay for, instead of the codec you don't have to pay for, because Something Bad Could Happen if you don't." Being vague and threatening is pure win, especially if you don't think you could actually prevail in court.
posted by Malor at 9:27 AM on January 14, 2011


(that last bit isn't pointed at you, jedicus, I'm talking about THEM being vague and threatening. I don't particularly agree with you, but you don't strike me as being deceptive in your arguments.)
posted by Malor at 9:30 AM on January 14, 2011


that last bit isn't pointed at you, jedicus,

Fair enough, but it would've worked equally well to simply leave my name out of it. And it's not like Google doesn't have plenty of folks arguing its case for free as well.

In any case, I don't think any real decision makers are going to be swayed by discussions on the internet, and end users really don't have anything to fear, so this discussion shouldn't really affect anybody's choices. Nobody is going to sue end users for patent infringement. It would be a massive money-loser (seriously we're talking the reasonable royalty from each end user amounting to perhaps $1 at most; the filing costs alone would be hundreds of times that), and a PR disaster if it came from a company that actually sold stuff.
posted by jedicus at 9:58 AM on January 14, 2011


The 'no-takebacks' license is nice, but it doesn't legally preclude Google from tweaking WebM and releasing it with a more-restrictive license. I admit it's splitting hairs to some degree to say this, but all that you get from this license is the current version, and they're very explicit about the fact that future revisions might have different licenses. Like, say, they tweak the implementation so that it is easier to hardware-accelerate, or uses 50% less CPU, or fixes a security hole, or whatever, and charge royalties for using that.

There was actually a kind-of strategy based on this concept at a company I worked for. Open-source the product, build momentum behind it, then make compelling improvements to it while closing the source. In software, the updates matter as much or more than as the current version, so this would eventually suck everyone in to using the proprietary version. Of course such a move would generate a tidal wave of hatred, but if you had a monopoly the hatred wouldn't matter.

Anyway, all this is to say is that Google is not being evil, they are simply reserving the right to potentially do evil in the future, and I think it's reasonable to be more worried about a single entity doing so than a consortium of competitors doing so.
posted by breath at 11:56 AM on January 14, 2011


Oh, I'm not disputing that
a) there are software patents granted at the national level in european or other countries, either because they were granted on something that is patentable (i.e. hardware), or the relevant patent office cocked it up.


Nope, software is patentable in Europe, what isn't patentable are "computer programs as such". It has always been possible to patent not just physical products/devices, but also processes/methods (for instance, a new method for synthethizing a known chemical compound, or for controlling an existing device). A new and inventive technical process is thus perfectly patentable, whether it is carried out using hydraulic valves, a hardwired circuit, or a generic CPU running a program.

Video compression is a technical process, and therefore patentable. Filling your tax form isn't a technical process and therefore tax software won't be patentable in Europe. In between there's a murky, grey area which lawmakers are terrified of addressing because each time that the subject is mentioned they get a storm of angry e-mails from misinformed geeks.

However, it has been tested repeated in european courts that the EPO standard applies; they are not enforceable. France banned software patents in 1968, for example.

Any patent action taken by MPEG-LA against webM users will not take place in europe. Well, unless they want to play patent troll, and just try and outspend companies with a case that is unwinnable. European courts are a lot less tolerant of that than US courts though.


As I said, the EPO "standard" (Arts. 52(2) and (3) of the European Patent Convention) actually allows that kind of patent, and indeed most of those MPEG-LA patents will have been granted by the EPO itself. Moreover, courts have upheld such "software patents" all over Europe (though they still can't agree on where to draw the line: for instance, English courts are generally less friendly than German courts towards patents that explicitly mention a computer program). Indeed, since the Bilski case, and apart from some additional murkiness in the US with respect to business methods, the situation in Europe and in the US is not hugely different anymore.

Moreover, some European courts are, if anything, more patentee-friendly than US ones. While you won't find the same kind of damage awards in Europe than over the pond, litigation, on the other hand, is one order of magnitude cheaper and German courts, for instance, are particularly injunction-happy, something which Sisvel, which manages a patent pool for MP3 (software!) has already made good use of in the past.

Mathematics alone - i.e. codecs - are not patentable.

But codecs aren't "mathematics alone" if they result in a video being played on my screen. That's why they are more than "computer programs as such".

Look, neither jedicus nor me are shills for MPEG-LA. But we both know the patent system and we both know that it is ridiculous to suggest that whereas H.264 is "patent-encumbered", WebM isn't. Moreover, H.264 is an ISO/IEC standard, and ISO has very strict rules forcing any company taking part in the definition of a new standard to disclose any related patent or patent application and to make it available under FRAND (Fair, Reasonable and Non-Discriminatory) terms, and that's where all the patents managed by MPEG-LA come from. WebM hasn't gone through the same kind of scrutiny.

And, as jedicus points out, the companies contributing to the MPEG-LA pool have a strong interest in ensuring that their licensees aren't hampered by other patents, whereas, as the smartphone patent wars (with several battles being played out also in European courts) are showing, Google doesn't care much if their licensees get pounded with two dozen lawsuits.
posted by Skeptic at 1:34 PM on January 14, 2011 [1 favorite]


Great move by Google. We had to encode in WebM in the coming months anyway since Firefox, which has a large user base, doesn't and would never be able to support H.264. Chrome is said to have a higher user base than Safari but still currently around the 10% mark (and growing) so its great to see Webkit get this support early on as webkit browsers continue to grow in use.

Flash support is there for older browsers and operating systems as a fallback (and for animation until Canvas is widely used, developed, and mature) and this is why YouTube will continue to serve in multiple formats for years (though it was nice of them to say bugger to IE 6).

As for the patent-encumbered/not encumbered bit, yes, there is a potential for lawsuits but as of today, WebM can be used free of charge and distributed as well. As far as I know, XHTML isn't an ISO/IEC standard as well, but I don't see people worrying about potential lawsuits over markup languages so let's just go with something that we can pay for instead.
posted by juiceCake at 9:16 PM on January 14, 2011


I wouldn't much care about the H264 vs WebM fight except that it helps bring Flash back from its impending doom. And that sucks. Flash needs to die.

This doesn't affect Flash much at all. And until systems are fast and efficient enough to handle Canvas (or SVG, should somehow that prevail over Canvas in a late upset), Flash will be the primary choice for motion graphics. HTML5 isn't even a standard yet.

Hell, at least with Flash you can stream. As the standard is currently written, you can't stream with the video tag (though that will get fixed at some point in the future).

Flash on the other hands burns CPU (and therefore battery life) like a bitch.

Next to these larger implementations of Canvas I've seen, Flash is a flippin' Prius.

I know there's a lot of DIE FLASH DIE DIE DIE in the community, but I think most of us who build websites for a living know that isn't possible, not with the web as it currently is.
posted by dw at 11:22 PM on January 14, 2011 [1 favorite]




Where the choice is not being able to view the standard codec at all in Safari because Apple is insisting on h264, or being able to view it through a plugin, I fail to see how Google's plugin is anything but promoting the standard. The idea here being that with internet video running on WebM, and users either using a browser that supports it or a plugin for browsers that don't, maybe Microsoft and Apple will get their heads straight and support it themselves instead of leaving their users to get core functionality from a third party.
posted by kafziel at 7:54 AM on January 15, 2011


Chrome users are the latest casualty in Google's crusade against Apple
Second, If Google really wants to promote open technologies, they why is it so cozy with Adobe and Flash? For most definitions of “open”, HTML5 and CSS3 are open but Flash is not. Think of all the resources Google put in to sandboxing Flash and checking for Flash updates. Google could use those resources towards making HTML5 more functional and making Flash and other plug-ins obsolete. Ah, but Apple doesn’t support Flash, and there’s the rub.

Google can use Flash as a foil against Apple because Android supports Flash but iOS does not. Google’s double standard is especially noticeable if you realize that Flash supports H.264 video. So Chrome will still play H.264 video, but only if you use a Flash plug-in instead of the standard <video> tag. Clearly, it’s not about openness or cost savings: this codec announcement is just another political jab at Apple.
posted by Blazecock Pileon at 10:18 AM on January 15, 2011


Where the choice is not being able to view the standard codec at all in Safari because Apple is insisting on h264, or being able to view it through a plugin, I fail to see how Google's plugin is anything but promoting the standard.
This choice of language, "the standard," is at the very least controversial, but more fairly would be described as outright false. There are many video standards, and if one of them is going to be called "the" standard, it's the one that has support of hundreds of companies and projects and is broadly deployed with many interoperating implementations both in hardware and software.

WebM is a developing standard, but the key part is "developing." WebM has very few software implementations (only one?), and no shipping hardware implementations. The battle for 2011's "standard" video technology was fought 8-5 years ago, and H.264 won. Right now, the battle being fought now is for "the standard" in 2016 and beyond. I hope, and am extremely confident, that Google's enormous corporate weight will develop WebM to the point that it's at least the technological equal of H.264. I hope that some day WebM will be sheltered by many different competing interests, rather than a single entity that also happens to have a great deal of control over where I go on the Web. If the days arrives where these two things happen, I hope and am confident that WebM will become "the standard." But that day is not today, and Google deciding to remove the standard video technology of today is an extremely heavy-handed maneuver that puzzles and concerns me greatly.

WebM is owned and controlled by a single enormously powerful corporation that seems to have good intentions. H.264 was created by an entire industry of cut-throat competitors so that they could work together, and possibly to keep out any new competitors. As much as this is a battle about patents, it's also a battle about accepting a benevolent dictator and rejecting an entire industry's decision.
posted by Llama-Lime at 1:32 PM on January 15, 2011 [2 favorites]


I doubt there's much benevolence in Google's decision. The argument presented by Blazecock makes a lot of sense: as long as the ability to play Flash remains Android's main USP against iOS, Google will have a vested interest in sabotaging any initiative that may render Flash obsolete.
posted by Skeptic at 6:19 PM on January 15, 2011


Is the removal of H.264 from Chrome a step backward for openness?
Haavard:

"So the question of Google’s bundling of Flash is a red herring which takes away the focus from the real issue: Whether native video support in browsers is based on open or closed technologies."

The problem is that it isn’t a red herring. It’s just another, actually larger, issue which he’s sidestepping.

As Google themselves wrote two days ago (bolding mine):

"Though H.264 plays an important role in video, as our goal is to enable open innovation, support for the codec will be removed and our resources directed towards completely open codec technologies."

How can that be your stated goal and you still bundle Flash? If you really want to enable “open innovations”, remove H.264 AND Flash from the browser.

Why is “open” allowed to be selectively applied to certain things and not others? Isn’t that what Google always gives Apple shit for?

It’s totally hypocritical. But the truth, as we all know, is that the “open” argument here is actually the red herring. Reading this, I’m not sure the people at Opera actually understand that, though.

posted by Blazecock Pileon at 3:25 AM on January 16, 2011


Chromium blog: (Chromium is the open source version of chrome)

"Some have expressed concern that our announcement will force publishers and developers to maintain multiple copies of their content when they otherwise would not have had to," he said. "Remember, Firefox and Opera have never supported H.264 due to its licensing requirements, they both support WebM and Ogg Theora.

"Therefore, unless publishers and developers using the HTML video tag don’t plan to support the large portion of the desktop and mobile web that use these browsers, they will have to support a format other than H.264 anyway (which is why we are working to establish a baseline codec for HTML video)."

"It is clear that there will not be agreement to specify H.264 as the baseline codec in the HTML video standard due to its licensing requirements. Furthermore, we genuinely believe that core web technologies need to be open and community developed to enable the same great innovation that has brought the web to where it is today. These facts led us to join the efforts of the web community and invest in an open alternative, WebM."


This is the list of supporters of webM. There's quite a few big names in there, including on the hardware list - qualcomm, arm, broadcom, freescale, amd, nvidia, marvell, logitech, chips&media for example.

WebM has very few software implementations (only one?), and no shipping hardware implementations.

WebM is included in ffmpeg via ffvp8, an alternative implementation to google's, and since so many other projects draw on ffmpeg for codec support that's a lot of other projects that get VP8 support.

Chips&Media have a hardware video chip that includes h264 and VP8 1080p 60fps playback. The nexus S android phone with samsung ARM has hardware webM support.

---

Google has made it quite clear that it needs Flash because YouTube needs Flash. YouTube can't rely exclusively on HTML5 until the standard is finished properly (which it isn't yet - the spec around streaming is muddy). What's more, Flash is an advertising tech.

Eventually, Google prefers a web - including YouTube - that handles all video with HTML and an open codec. That best suits the company's bottom line, which rises as general use rises, and ultimately, use rises with free and open standards. But in the short term, its bottom line would suffer if - in an effort to hasten the march towards this ideal world - YouTube suddenly went Flash-free.

Netflix have stated that there are a number of issues preventing them from using HTML5 video: acceptable A/V container formats; acceptable audio and video codecs; streaming protocol; a way for the streaming protocol to adapt to available bandwidth; a way of conveying information about available streams and other parameters to the streaming player module; a way of supporting protected content; and a way of exposing all this functionality into HTML5.

The fight over what will be standard codec for native video streaming, without plugins, is far from over. I don't think either side will win - apple and microsoft gain too much from locking out open source browsers, and google, mozilla, kde and opera all make open source browsers - and given each side has roughly 50% of the browser market, that's not going to change anytime soon.
posted by ArkhanJG at 4:23 AM on January 16, 2011 [1 favorite]


Oops, opera shouldn't be included in the open source browser list, but they are on the same side with webM only support.
posted by ArkhanJG at 4:26 AM on January 16, 2011


"Therefore, unless publishers and developers using the HTML video tag don’t plan to support the large portion of the desktop and mobile web that use these browsers, they will have to support a format other than H.264 anyway (which is why we are working to establish a baseline codec for HTML video)."

The baseline was H.264. Google is trying to change what was pretty much settled on by everyone, except the Mozilla Foundation.
posted by Blazecock Pileon at 4:34 AM on January 16, 2011


And opera and konqueror, as well as mozilla. Who make up some 25-30% of the browser market (and growing), and mozilla simply cannot support h264 due to their licence. That's not a small section of the market. Opera is also huge in the mobile browser space.

h264 was never a baseline for everyone, only for the proprietary segment of the market. Even in that segment, it's not actually deployed in any size yet.
posted by ArkhanJG at 4:47 AM on January 16, 2011


Opera is proprietary, and Google is supporting a proprietary solution by — deliberately or otherwise — pushing Flash. Google uses Apple's open source WebKit to power its web browser. Upon a deeper examination of the landscape of all the bits and pieces, it just doesn't seem like notions of "open" and "closed" enter into Google's true motivations about any of this.
posted by Blazecock Pileon at 4:56 AM on January 16, 2011


And webkit is apple's minimal fork of KHTML, from KDE - an open source browser component on linux. Which is why they had to make it LGPL too. So your point is that open source code is good, and makes life easier to produce new products without reinventing the wheel every time?

Google also massively fund the mozilla foundation, an open source browser. They've released WebM and VP8 as an open source product, free to use and implement by anyone - as the ffmpeg project have done. They're working an android, an open source OS based in large part on linux. They're also working on ChromeOS, another linux variant for desktop/netbook use.

Google benefits massively from open standards that anyone can use, not just microsoft in windows, and apple in OSX. They need and want an open platform for everyone, not just those with the cash who buy windows or a mac. After all, the more people they get using their services, on whatever platform, the more eyeballs they have for their brokered advertising, their real money spinner.

Opera also consistently support open standards, if not open source, so usually end up on the same side as the mozilla foundation, google, kde et al. They have specifically stated they don't support h264, never have, and originally pushed for the video tag, using ogg theora as it was the only significant open video codec available to all at the time. Theora is pretty crap, admittedly, but VP8 is much better, roughly on a par with baseline h264 - and it will only improve in time.

Google want to make money, no question, thus the need to keep flash support for youtube. Flash is available on linux, in firefox no less. It is the current standard, whether we like it or not. I don't like flash/h264 either, but replacing it with h264 native is *just as bad*. This is about what happens for the rest of this decade. Whether its only for guys with big pockets or not.

The only people who don't support flash are apple on iOS. How is pushing HTML5 with h264 before its ready as 'the' web standard anything other than an attempt by apple to cut everyone that's not apple off at the knees, based on their success with iOS? To force the issue and cut 25% at least of the web users off from the future of web streaming unless they bite the bullet and also switch to a proprietary, patent encumbered solution?

So much of what google produce and support is open source as well as using open standards, so by throwing their lot in with mozilla et al with chrome there's a greater chance of people NOT running IE or safari still having open access to the internet in the next decade. How is that a bad thing?
posted by ArkhanJG at 5:28 AM on January 16, 2011


webkit is apple's minimal fork of KHTML

Calling it "minimal" is a bit of a mischaracterization. Google didn't make their own engine or use KHTML. They picked WebKit. So "minimal" is either wrong or suggests Google doesn't have the chops to make competent technical decisions.

So your point is that open source code is good, and makes life easier to produce new products without reinventing the wheel every time?

No, my point is that even Apple releases Real Open Source software that even Google will humbly use in its own products. Not everything Apple does is evil or hurts everyone else, and, in fact, WebKit is a substantial work in its own right, one that has benefitted many other parties, including Google.

Google also massively fund the mozilla foundation

In exchange for the Mozilla Foundation contractually directing web search traffic to Google with its products, not because Google has an agenda of promoting open source, per se.

VP8 is much better, roughly on a par with baseline h264

Compared with H.264, VP8 apparently has technical weaknesses that are locked into its specification by Google. Probably only its proponents would put VP8 on par with H.264.

The only people who don't support flash are apple on iOS.

If the goal is to push VP8 for the reason that it is an open standard, Google would have dropped Flash support, as well. They chose not to, but instead supported both VP8 and Flash.

My guess, and one shared by a number of others who have looked into this, is that Google has a long-term vision of putting the screws to Apple, by whatever means at hand. Whatever the reason, it doesn't appear that promoting Open Source as a movement has a whole lot to do with it.
posted by Blazecock Pileon at 6:21 AM on January 16, 2011


They picked WebKit. So "minimal" is either wrong or suggests Google doesn't have the chops to make competent technical decisions.

webkit is still very, very similar to KHTML. It is however ported to windows, for use in windows safari. KHTML is aimed for KDE and QT, on linux, unsurprisingly as it's a KDE project. Work from webkit has been backported into KHTML, but some hasn't as it's specific to OSX.

It makes sense to use that existing work under LGPL, rather than reinvent the wheel. Apple does the same with parts of freebsd/openbsd for OSX. Hell, even windows has BSD code in it.

That's rather the point of open source code - sharing and reuse. Apple use KHTML to make webkit, google use webkit to make chrome. Chrome is the basis for chromium on linux, and the circle is complete. Kudos to both of them. But to spin webkit as some wonderful feelgood open source project of apple that google is sponging off of rather misses the point - under that view, apple is sponging off of the work of KDE just as much, and have to release their source as they based it in large part of someone else's work, the licence of which requires such changes to be released also.

Thing is, that's what open source is for. Neither is sponging, they're doing what the original code writers wanted - take good code, make it fit their own use, and release their modifications for anyone else to use too.

Of course google has the long term view of 'screwing' apple. Apple wants a closed ecosystem where everyone only uses apple products. Google wants a system where everyone can access google services, and see their advertising. Both are diametrically opposed. Neither of them are 'good guys' here, nobody is in the tech world. It just so happens that an open infrastructure and open standards improve the web for all users, not just one company. That's the whole point. Even if someone works on open source code for entirely the wrong reasons it doesn't matter in the slightest - the code is still open, and out there for everyone.

Microsoft retarded the development of the web as a platform for many years with IE and its breaking of standards using its dominant OS position, because it suited their goal of making everyone tied to their platform.

Tieing HTML5 to h264 does exactly the same thing - split the web between those that can implement it, and those that can't. I don't want a future that's dependant upon any one company, be it abobe, apple or google.

Yes, I hope that google drops flash. I think they should drop embedded flash from chrome. But they do that to make the *current* web easier to use, and get people away from IE. VP8 is a standard that everyone can use, not just those who can afford the patent fees and provide proprietary software.

I wondered how long before that blog entry would show up. It's from when VP8 was initially released, and was a straight code dump of the On2's work. The spec has undergone revision in the last 6 months, along with much work on the code itself. FFmpeg have a significantly improved implementation in ffvp8 - the same developer was involved in making that ffvp8 version.

Is h264 high profile better? Yes, significantly. It's aimed at bluray levels of rendering, an entirely different beast than web streaming. I did see a review of ffvp8 vs x264 baseline the other day that concluded it was basically a wash in quality with each having different strengths, but I can't find it and I have other things to get on with today. I've found this one though.

If I manage to find the other review later, I'll stick it in here.
posted by ArkhanJG at 6:57 AM on January 16, 2011


The baseline was H.264. Google is trying to change what was pretty much settled on by everyone, except the Mozilla Foundation.

by 'everyone', you mean apple. Since obviously Google, Firefox and Opera weren't on board.
posted by empath at 7:10 AM on January 16, 2011


Personally, I'm not sure I care if Google is promoting an actually free and open alternative to h264 because it will help Larry, Sergei, and Eric finally realize their desire to consume kitten-babyseal smoothies produced by sweatshop labor. I am completely sure, however, I'm not going to care at all if Apple's influence suffers a bit. The issue of having any significant piece of a web standard not be freely re-implementable is way more important.

This choice of language, "the standard," is at the very least controversial, but more fairly would be described as outright false. There are many video standards, and if one of them is going to be called "the" standard, it's the one that has support of hundreds of companies and projects and is broadly deployed with many interoperating implementations both in hardware and software.

h264 may be a standard, but that's beside the point. It's not a web standard until anybody can sit down, write, and widely distribute authoring or rendering software for it without consulting or paying anybody else fees for the privilege.

Google may or may not have various ulterior motives, WebM may or may not be technically on par with h264, WebM may or may not be attacked with patent claims. But WebM meets that important criteria for a web standard. Right now h264 can't, just like Flash can't. Though at least Flash isn't pretending (despite arguably passing the test I'm talking about better than h264 does).

How can [enable open innovation] be your stated stated goal and you still bundle Flash?

If we're going to get into making value judgements of a given party based on whether they do anything to support any proprietary technology that overlaps and therefore competes with web standards no matter what their other contributions are, Apple is not going fare well.

I'd prefer to think that reasonable people can see that driving towards a freely re-implementable standard baseline for the web and having some measure of support or even enthusiasm for things beyond it are two separate issues.

My guess, and one shared by a number of others who have looked into this, is that Google has a long-term vision of putting the screws to Apple, by whatever means at hand

By doing what? Putting WebM out there in order to promote Flash in order to make Apple suffer? Doesn't make any sense at all.

* It's a move Apple can easily sidestep by simply... supporting WebM.

* In fact, Google will help Apple circumvent their ostensibly evil plan to boost Flash by releasing plugins and code and license grants that would enable them to support WebM with minimal effort and negligible cost

* The members of the MPEG-LA pool could also sidestep this by freely granting a license to implement h264 in software (as well as their claim on licenses for endpoint media sales and other shenanigans), at which point if it fares well by technical standards, I suspect everybody would embrace it, and they could probably still do pretty well with their hardware licenses.

* Despite Flash's important role in web video, the idea that it's the reason for Flash's success is a misconception. It grew to ubiquity almost entirely on other applications, and even if the video codec brouhaha was resolved today, Flash isn't going away until browsers all have standard sockets, easy streaming, full duplex media communication, vector graphics, bitmap graphics, and audio (including programatically generated audio, which as far as I can tell only the Mozilla people are even thinking about), and probably some other things I'm forgetting. Oh yeah, and there's point-and-click authoring tools for all this (probably written by Adobe).

As a plan enshrine Flash (which would cause Apple to suffer why again?), it's full of enough holes that Google would have to be pretty dumb to think this tack is going to "put the screws" to Apple.
posted by weston at 12:16 PM on January 16, 2011


Blazecock, on the assumption that you're a typical end user, I think you're arguing against your own interests.

If H264 ends up becoming 'the video standard', that means that Firefox will forever be a second-class citizen, as will any other open source browser that might arise. Either they stay absolutely marginal and unimportant (sub 100K copies shipped), or else are forced to monetize their user base in some way to pay for licensing the codec.

This means that the entire open source ecosystem can't integrate video into other products. If H264 is the standard, the whole free software movement is effectively shut out from ever using it. They'd have to work in some other free codec, and then transcode to H264 for presentation, which always impairs quality. Further, that final step would cost money, if it was being done by a piece of software that was at all common.

Yes, Google is advancing its own interests in these moves, but its interests align much more closely with the average user's than pretty much any other large corporation. An open, standards-based web lets them thrive, but it lets OTHER entities thrive too, able to innovate and create new products without an imposed tax. Google is explicitly trying to REMOVE handcuffs from end-users and developers, and MPEG-LA is specifically about putting them on, extracting a toll from pretty much everyone of significance.

Google thrives when its customers thrive. Yes, refusing to support H264 is good for them, but it's good for you too. So far, at least, Google's relationship with the public is largely symbiotic, not parasitic. The H264 licensing standards are most certainly parasitic.

And nobody really gets hurt by this, except for a few companies that wanted to charge toll. Even web developers benefit. With Google's half-plugins for Safari and IE9, you'll be able to encode to WebM, and present it to any client using a completely standard HTML5 interface. How that client gets the ability to render the file becomes irrelevant to you. You don't have to see or deal with the plugin, you serve up the video. If you wish, you can encode to both H264 and WebM, but you'll be able to stick with just WebM and be universally compatible.

If you planned to standardize on H264, you were going to be shutting out about 30% of the browser market, since Firefox can't support a payware codec. So you would either have had to tell users to use a different browser, or encode your files twice. In the WebM scenario, you encode just once, and everyone can use it. You can also encode to H264 if you wish, but it's not necessary.

Old way: either tell 30% of your users to piss off, or encode to two standards.

New way: Encode once. If you wish, you can encode to more.

The ONLY people losing out here are the toll extractors.
posted by Malor at 3:27 PM on January 17, 2011 [2 favorites]


I had some more thoughts today, after reading some threads elsewhere.

It is true that iPad and iPhone users won't get as good an experience with WebM, because the hardware on those units only accelerates H264. But optimizing HTML5 video for a specific subset of devices that are shipping in 2011 is going to look awfully silly in 2021. Most clients will still be able to render WebM, even without acceleration, and standardizing on something that happened to play well over a very short technical window means that we'll be stuck with that format essentially forever. Apple and iDevice users get a big short-term advantage, at an imposed cost to all video on the web for the foreseeable future. That's a really bad idea.

We are making ourselves somewhat beholden to Google in accepting this standard, without a doubt. But the way they've released the code and promised not to sue on their existing patents means that, even if they do suddenly go evil, we can still keep the existing format and use it as long as we like. Google has explicitly made it impossible for it to take the released code and patents away in the future. They could implement NEW versions that require a license fee, but the existing version is free in all senses, and that won't change.

Contrast that with H264, which is definitely not free even in the present. And those already suboptimal terms are guaranteed only through 2015, when the license fees could be completely changed, making it even more difficult for small players to compete with the big boys. It's bad now, and we could potentially lose what rights we already have in four years.

With WebM, we have what Google is giving us forever... no matter what they do in 2015 or later, the 2011 version will still be usable, for free, by everyone. This is a much better deal. That's the kind of foundation you can safely build a business on, and H264 just isn't.

It is a shame that iPad and iPhone users won't get hardware acceleration. If Apple could wring as good a deal out of MPEG-LA that Google is giving us, and Lord knows they have more money than God, then iDevice users would be in much better shape, because the Web could safely standardize on H264 instead. But as is, it just can't, and if Apple customers are upset, they should be blaming Apple for choosing a closed standard for hardware acceleration, instead of finding or creating an open one. The fact that Apple was willing to sell their customers down the river by binding them to a non-free standard should make them angry at Apple, not Google.

Apple has gained a gigantic benefit from free software; it's the foundation of every OS they sell, top to bottom. Trying to impose a closed ecosystem and closed video standards on top of that free code strikes me as extremely unethical. As a direct result, their customers are, apparently, going to be badly inconvenienced.

This does not particularly upset me.
posted by Malor at 8:28 AM on January 18, 2011


Yes, Google is advancing its own interests in these moves, but its interests align much more closely with the average user's than pretty much any other large corporation.

This is the single most incredible statement I think I've read in a Google (or Apple vs. Google) thread on Metafilter. Either that or I'm the least average user ever, which in the population of Metafilter, I sincerely doubt.
posted by immlass at 9:20 AM on January 18, 2011


Blazecock, on the assumption that you're a typical end user, I think you're arguing against your own interests.

He's not. He's an apple partisan.
posted by empath at 9:24 AM on January 18, 2011


immlass, what I've seen of Google has consistently been aligned with the furtherance of open standards. Further, they seem very good about providing 'escape hatches', ways of extracting your data and moving it to competitors. Then they try to provide such good service that you won't want to.

How do you think Google's actions don't support your long-term interests?
posted by Malor at 9:33 AM on January 18, 2011


How do you think Google's actions don't support your long-term interests?

Man, I love Google. I mean, love them. I would have their babies if I could.

But they have a frightening amount of power and are accumulating more, and I do worry about how that's going to shake out if the management ever changes.
posted by empath at 9:45 AM on January 18, 2011


Well, you know, because they're so good at offering data extraction and opt-outs, I'm not sure they're that much of a threat, at least if they do it in any kind of visible way. They have, apparently purposely, largely crippled their own ability to successfully go evil.

I don't even use many of their services; I prefer providing my own infrastructure, because I know how, and don't like giving my data to any corporation. But I haven't seen anything from Google that would be terribly difficult to extricate myself from. I'm not seeing any obvious ways they could screw me, because I can switch over to a competitor fairly easily.

Maybe it's just a failure of my imagination, but I see Apple and Microsoft actively trying to screw me every damn day, right now in the present, where any potential threat from Google seems remote, vague, and improbable.

I see this instance as being another case of the same thing; the way they've structured this deal, there's no reneging. We get to use WebM, in its present form, permanently. There's some chance of submarine patents, but that's a theoretical threat, contrasted with the real and concrete ones that are already visible in H264.

Using H264 means that, in 2015, MPEG-LA has the world over a barrel. Using WebM means that, sometime in the future, Google can offer us a barrel, but the choice of whether to bend over remains purely with us.
posted by Malor at 10:17 AM on January 18, 2011


How do you think Google's actions don't support your long-term interests?

Google is an advertising company. They used to be a search company, but we're currently having a discussion elsewhere on the blue about how they're basically losing search under the sludge of SEO and are either incapable of using their vast resources to solve their problems or can't be bothered to. But now they make their nut on ads, and as a consumer that makes them Not My Friend.

I am not Google's user. I am Google's product to be sold to advertisers.

In this particular case, the technical and legal merits between WebM and H264 are not something I feel qualified to comment on: the technical side is over my head, and the legal theories about patent liability are meaningless until tried in court. And I, personally, don't give a damn about the nitty gritty of the politics of which version open is actually the right kind of open. But where I can tell Google is a bunch of hypocrites is when they argue "open standards" when it's time to stick it to H264 (and their competitor, Apple) and "market share" when it's time to keep Flash. Of course, the benefit to Google about Flash is that it's a platform that advertisers like; keeping it around has nothing to do with average users and it is not particularly in the interest of average users to keep it.

This may put a nail in the coffin of H264 and it may be better for the internet overall (infrastructure, "content providers", and end-users collectively) to do so, if only because picking a standard is better, but it has nothing to do with users, unless you mean Google's paying clients (advertisers).
posted by immlass at 1:52 PM on January 18, 2011


But now they make their nut on ads, and as a consumer that makes them Not My Friend.

I am not Google's user. I am Google's product to be sold to advertisers.


I can't even disagree with you there all that strongly, as far as being a product goes. But, at least so far, they seem to be very, very good about both advertising to you in 'fair' ways, typically with relatively minor obtrusiveness, and in realizing that abusing you will result in your defection to the competition. And, this is particularly important, they've been very careful to leave that option open to you in the products I've personally looked at. That's very, very rare in the corporate world. Apple doesn't give you any easy way to opt out of the App Store, for instance; their assertion of continued control over a general-purpose computing device harkens back to the bad, bad old days of mainframes. Microsoft seems to have kind of lost focus these days, but there was a time when they were an actively malignant force in the computing world. And don't even get me started on Facebook.

I think of a company as 'evil' when they hold you arbitrarily captive, or share data in ways they shouldn't, and I've never seen Google doing that. They do have a lot of power, but it's very 'soft' power... if they actually tried to use it to truly exert control, it would evaporate. About all they're able to do, in most cases, is make suggestions.

In this particular case, they're not supporting a payware codec on their 8 or 10% market share browser. And they're following the lead of the 30% share browser, so it's not like the idea of abandoning H264 is particularly theirs. However, being richer than Croesus, they're in an unusually good position to offer a replacement, and they're doing so on extremely liberal terms, so liberal that they'll never be able to take it back.

Whatever you think of Google's true motives, this is a good deal for almost everyone involved with the Web. Because of the way they've structured the code and patent license, it appears very unlikely that they can pull a fast one on us. This is very, VERY untrue of H264.

But where I can tell Google is a bunch of hypocrites is when they argue "open standards" when it's time to stick it to H264 (and their competitor, Apple) and "market share" when it's time to keep Flash.

I don't think 'the preservation of Flash' has really been much of a driver for them. Using Flash for video wasn't likely to go away anyway, because there's no DRM in the HTML 5 video standard, at least not yet. There may never be. Companies like Netflix are contractually unable to stream video content if they can't protect it, so unless and until Hollywood starts to get a clue, we're probably going to have Flash for most commercial video. And this would happen no matter what codec Google supported.

In other words, even if they totally DID back H264, it wouldn't make any particular difference to the continued existence of Flash on the web. That argument is a red herring, trying to get you to take your eyes off the ball. Much of the reason Flash is so pervasive is because of the lack of a video standard, but ANY video standard will get rid of that reason to use it. The DRM and presentation issues will remain in any case. If you're a Flash-hater, that's fine, but don't think that you OR Google supporting H264 over WebM would have any particular impact.

Of course, the benefit to Google about Flash is that it's a platform that advertisers like; keeping it around has nothing to do with average users and it is not particularly in the interest of average users to keep it.

There's some merit to that observation, but again, this particular decision really won't affect it very much. Google will be providing "half-plugins" (invisible to the server) for at least Safari and IE9, so providers that want to junk Flash and go to HTML5 video will mostly be able to do so seamlessly, without having to think about or care what's installed on the target system. Those that want to directly support the iDevices may have to encode twice, because Apple may refuse to allow the WebM plugin on its closed platform. But that's Apple being evil, not Google. Battery life will suffer on those systems, but settling on a permanent standard, so that a small subset of the total video-using population can go longer between charges, wouldn't be very smart. They'll be replacing those devices within a couple of years anyway, most likely. Standards last a long, long time... individual generations of hardware don't.

Google is also designing hardware accelerators for WebM for mobile devices. According to comments on the lwn.net site, we should see actual shipping consumer hardware with the accelerator by the end of this year. So it's a temporary problem.

This may put a nail in the coffin of H264 and it may be better for the internet overall (infrastructure, "content providers", and end-users collectively) to do so, if only because picking a standard is better, but it has nothing to do with users, unless you mean Google's paying clients (advertisers)

I strongly disagree with that, because being better for the Internet overall IS better for users. You seem to think that it's an accidental byproduct, but are aware that it will end up being good for almost everyone. Even if it is accidental, we still get the same result, so Google's true motivations are kind of irrelevant. But I think you're badly misreading those motivations... Google is doing this to keep innovation going. Yes, they benefit from that, because they can buy up successful companies with great technologies. But you win too, with the great new ideas and technologies the next batch of startups will come up with. They win, you win, new startups win. What's not to like?

Finally, I want to go back to an early sentence:

And I, personally, don't give a damn about the nitty gritty of the politics of which version open is actually the right kind of open

I can't tell you to care about something you don't care about, but it's worth sorting through the various claims in this instance. It's becoming quite common for companies to claim 'openness' when they're not doing that in any meaningful way. H264 is a great example, because it's only free until you ship 100K copies, and that's explicitly only good through 2015... after that, they can impose any terms they like, because of their patents. That's an amazingly shitty deal, and is being rejected resoundingly by truly open software like Firefox.

In general, I think these definitions are one of the things worth paying attention to on the Web. If you only use commercial software, you may never care, but I use a great deal of free software, and sussing out what people actually mean when they claim "it's open!" is very important to me. Even if you're not using free software now, you may want to in the future, so learning a little bit about the various flavors of open, lowercase free, and uppercase Free might be worth your time. It's not particularly difficult. The only reason it sounds tedious and complex is because of companies misrepresenting their products, trying hard to confuse the issues so you'll buy in.
posted by Malor at 3:22 PM on January 18, 2011


I am not Google's user. I am Google's product to be sold to advertisers.

Of course, if we don't keep our users happy, we won't have anything to sell (in your analogy/statement/whatever). People didn't start using Google over other search engines because we're better at monetizing or whatever, they did it because they thought it was a better search engine.
posted by wildcrdj at 3:39 PM on January 18, 2011


Normally I don't give in to the impulse to reply point by point, but I'm going to do it this once because I think there are so many disagreements on basic premises that I should probably unpack them so people can see the points of disagreement.

I think of a company as 'evil' when they hold you arbitrarily captive, or share data in ways they shouldn't, and I've never seen Google doing that

I take it you've forgotten the (ideological) privacy issues around Google Buzz.

However, being richer than Croesus, they're in an unusually good position to offer a replacement, and they're doing so on extremely liberal terms, so liberal that they'll never be able to take it back.

Google being richer than Croesus is actually a reason to distrust it on both technical and user fronts. Because they're richer than Croesus, workgroups have no pressure on them to create quality products that end-users want, and no pressure to stop floating products that are half-baked/not ready for prime-time. Google Wave is a well-known example of this, but Google's history is littered with projects and technologies that seemed like a good idea, but have since been abandoned either for no apparent reason or for a lack of user uptake because they were unready or half-baked. Google is no more to be trusted to pick a winner or a better technical answer (or a more useful one, or one that will last in a moving internet landscape) than anyone else.

I strongly disagree with that, because being better for the Internet overall IS better for users.

I disagree with your premise. Benefits to infrastructure or content producers or hosts can be horrible for end-users. Is it good or bad to allow certain data types to be prioritized? That may depend on whether you're the network trying to manage bandwidth and traffic or the end user wanting your torrent/Netflix instant/whatever now.

Standards are overall a good thing but that doesn't mean this standard is better or that Google will pick a good one even if it's technically better. Flash and VHS/Beta and the sound quality of MP3s all come to mind.

The only reason it sounds tedious and complex is because of companies misrepresenting their products, trying hard to confuse the issues so you'll buy in.

I personally have found sorting through the claims of various ideologues much harder to untangle than sorting through what various corporations have to say. Apple and Google don't always believe their own PR but that's hardly true of open software activists I know.

The question on patents boils down to: who do I believe, Apple or Google? I suspect Apple and Google are both telling the truth as they see it. What they say will be factually correct as far as it goes, but shaded to match their agenda; they'll purposely neglect to mention facts that put their actions in a bad light. That's what big companies do.

Last but not least: they did it because they thought it was a better search engine.

Isn't this the discussion that we're having in the other thread? Google may still be "the best search engine" but I know I'm taking queries to my FB friends and AskMe over googling for things more and more often these days because Google's results are less and less useful. Best does not mean good.
posted by immlass at 4:35 PM on January 18, 2011


But where I can tell Google is a bunch of hypocrites is when they argue "open standards" when it's time to stick it to H264 (and their competitor, Apple) and "market share" when it's time to keep Flash.

I suspect that rather than hypocrisy, it's the ability to see the distinction between the implications of proprietary claims inside core web standards and competing (closed or open) technologies outside of it, not to mention that despite Flash's failings, there's a real argument to be made that it fares better than h264 on credible definitions of "open."

I can't read their minds and know if this is their actual motivation, but their actions are consistent with that of an actor who understands that the only acceptable web standard is one that's freely re-implementable, something we'll lose if h264 becomes the de facto codec for the <video> tag.
posted by weston at 5:30 PM on January 18, 2011


their actions are consistent with that of an actor who understands that the only acceptable web standard is one that's freely re-implementable

Their actions are also consistent with a mission of providing values to advertisers even at the expense of their nominal web-politics ideology. Chrome could have a Flash plug-in (in the same way Firefox does) but it comes with the browser. It's just apparently more important to have advertising baked into Chrome than it is to live up to the ideology of openness.

I like plenty of Google products but I'm not blind to the rule of huge corporations: follow the money.
posted by immlass at 8:18 PM on January 18, 2011


Their actions are also consistent with a mission of providing values to advertisers even at the expense of their nominal web-politics ideology.

Possibly true, but essentially irrelevant to my statement you're responding to and the issue at hand. Maybe it improves their position to take actions that will keep the actual web standard from becoming non-free. Like I've said earlier in the thread, that issue is important enough I don't really care much why they do it.

It's just apparently more important to have advertising baked into Chrome than it is to live up to the ideology of openness.

Why do you think that Flash is at all related to this issue? And how much of this thread have you read?
posted by weston at 8:56 PM on January 18, 2011


Why do you think that Flash is at all related to this issue? And how much of this thread have you read?

I've read the whole thing. It's about the same as most Google-related threads I read on the blue: lots of naive expression of faith in Google's "openness" as defined by Google, which tech people generally interpret as whichever particular brand of free they approve of most ideologically. If I distinguished this one in any particular way from the usual, it would be in reduced Apple-bashing because Blazecock Pileon is participating less.

Flash is only important in the sense that it puts the lie to "Google demands openness to bake it into Chrome" and exposes their double-standard. If you're happy with Google's decisions, whether it's because you support WebM for technical reasons, for ideological reasons, or because you want to see Google stick it to Steve Jobs, more power to you. I'm just bemused that tech folks actually trust Google to Not Be Evil when they're an ad company. The kool-aid of free software/open source/standards/etc. seems to cover a lot of sins.
posted by immlass at 10:10 PM on January 18, 2011


Google being richer than Croesus is actually a reason to distrust it on both technical and user fronts. Because they're richer than Croesus, workgroups have no pressure on them to create quality products that end-users want, and no pressure to stop floating products that are half-baked/not ready for prime-time.

You who else was "richer than Croesus?" Microsoft. In fact, they still are. And their mistake was that they did exactly what you suggest -- they skimped on quality and keeping the customer happy. What happened? Google ran them over.

Google Wave is a well-known example of this, but Google's history is littered with projects and technologies that seemed like a good idea, but have since been abandoned either for no apparent reason or for a lack of user uptake because they were unready or half-baked.

You talk like this is a bad thing. Google Wave itself wasn't a bad idea -- a single interface for a team to collect and curate and collaborate -- but it was way overhyped, and it was in some ways a solution looking for a problem. Buzz was a bust, but it does seem like if your name isn't Twitter or Facebook your social networking idea is going to suck (see Ping and all the various Microsoft disasters).

However, there are a ton of Google ideas that started as experiments that we now take for granted. Google is still trying and still having some successes amid all the weirdness that comes from Google Labs.

Google is no more to be trusted to pick a winner or a better technical answer (or a more useful one, or one that will last in a moving internet landscape) than anyone else.

I think we all agree to that. I just wonder why everyone freaking out about H.264 trust MPEG-LA -- and by extension Microsoft and Apple -- trust them more than Google. Because that just doesn't make sense.

Google has their motivations for WebM, but their decision to release the patent is ultimately beneficial to everyone. It also means Firefox won't have a premature end if MPEG-LA decides to come calling on the Mozilla Foundation for money.

And yet, it's all about how dare they take native support out of Chrome.

Flash is only important in the sense that it puts the lie to "Google demands openness to bake it into Chrome" and exposes their double-standard.

And for the thousandth time, the Flash/H.264 comparison is apples and oranges. Adobe and Google have a business deal for them; Google has never said Flash was open. WebM/VP8, though, is. H.264 is governed by a trap-door patent release. And functionally, Flash in Chrome is the same as in any browser -- a plug-in. It's listed on about:plugins, not about:credits (where all the credits are for open-source software used in Chrome).
posted by dw at 11:49 PM on January 18, 2011


I just wonder why everyone freaking out about H.264 trust MPEG-LA -- and by extension Microsoft and Apple -- trust them more than Google.

I trust any consortium more than any single standard holder because the consortium is likely to force the holders to play nice (which means in this case I'm not as concerned as many people in this thread about the idea that Firefox will get shafted somehow in 2015 or further in the future). Also, as much as I despise Microsoft and as much as Apple pisses me off, at least with them I know where the money is coming from and therefore who the customers whose interests are likely to win out are.

Fundamentally, though, you and others like you on the thread believe that Google is making the WebM decision on ideological grounds, and I fundamentally believe that the Chrome team has demonstrated that it runs on business logic, whether or not the WebM decision is right technically or ideologically. Someone upthread said something I found jaw-dropping, I commented to that effect, and now it's turned into a game of someone is WRONG on the INTERNET! I appreciate your comments and explanations, but we're coming at this from a fundamentally different paradigm about Google, and I've hit my "repeat myself three times" rule, so it's time for me to disengage from the thread.
posted by immlass at 7:33 AM on January 19, 2011


I trust any consortium more than any single standard holder because the consortium is likely to force the holders to play nice

As someone who's worked with the web for a decade plus, I can tell you that consortiums do no such thing. They're more like when the superheroes and the villains have to work together for a common purpose, like saving the world from a greater evil, or get a better rate on health insurance.

Choosing a consortium over a company is like choosing an oligarchy over a dictator. Either way, you don't have the power you think you do.

Fundamentally, though, you and others like you on the thread believe that Google is making the WebM decision on ideological grounds

No, I haven't said that, and I doubt anyone but a Google fanboi would say that. Google made a business decision to open WebM in order to benefit their long-term business model. But that decision actually benefits all of us, because WebM has no trap door patent issue. Google did this to make cash money, but there's no part of it that says if we use WebM we have to use Google products.

MPEG-LA's business model is to charge Apple/Microsoft/Adobe to use H.264, so the costs are passed on to us through higher prices on computers and software. Google's business model is to free WebM so they can widen the margin on YouTube ads.

In truth, the browsers probably shouldn't have ever had native H.264 OR WebM support. They should both be plug-ins. It aligns well with Microsoft "unbundling" decisions the courts pushed out in the 2000s.
posted by dw at 9:20 AM on January 19, 2011 [2 favorites]


I've read the whole thing. It's about the same as most Google-related threads I read on the blue: lots of naive expression of faith in Google's "openness" as defined by Google

I'm sorry, but no. I'm not going to let you even pretend that's an acceptable summary of the thread. For one thing, it's pointedly obtuse in its refusal to engage at least one objective definition of open that's been offered. If "freely re-implementable" is a definition of open you take issue with -- hey, even if you think think it would't matter at all to the web if HTML itself was "h264 open" and people who wrote HTML authoring or rendering tool might be subject to licensing fees and conditions by a consortium -- that's fine, but at least make some kind of argument on that front. But instead saying that this whole thread is just about people who like Google deciding to just take their word on some vague hand-wavy definition of Google because they like Google? Not OK. There has been some worthwhile argument about their un/trustworthiness, but there's enough argument beyond that on display here that if you really believe that's all there is to this, you have either not in fact read the thread, or you're suffering from a (hopefully temporary) reading comprehension problem.

Flash is only important in the sense that it puts the lie to "Google demands openness to bake it into Chrome"

I think this might be the fourth time I've pointed out that there isn't a contradiction between pushing hard for a freely re-implementable standards baseline and including functionality that goes beyond it. There's a whole conversation to be had here -- who knows, maybe with some kind of counterargument on your part -- if you develop any interest in processing and engaging it.

I trust any consortium more than any single standard holder because the consortium is likely to force the holders to play nice

You're speaking in probabilities and feelings about these entities when there are already outcomes. Google has already done everything you'd expect them to do for a freely re-implementable codec for the <video> tag. The consortium you're speaking of hasn't. "Trust" isn't really the issue at all.
posted by weston at 10:57 AM on January 19, 2011


I personally have found sorting through the claims of various ideologues much harder to untangle than sorting through what various corporations have to say. Apple and Google don't always believe their own PR but that's hardly true of open software activists I know.

You know, I gotta say, the sheer hostility in this paragraph is something I find mind-boggling. These guys and gals are cooperating to produce something they'll let you have for free. They correctly realize that software is hard to write, but free to copy, and so they share the task of creating the software and then let everyone have a copy, with the hope of getting further contributions to make it better still.

Of all the actors in the software world, these are the people you should be trusting the most. You use loaded terms like 'ideologue' and accuse them of lying, and explicitly say that Apple and Google are more honest.

You almost certainly take advantage of their code dozens or hundreds of times a day; it's freaking everywhere, from phones to routers to servers. And they ask no price; for the most part, their only terms are that if you take their code, improve it, and give that improved code to others, you have to release the improvements on the same terms. But your online voice just drips with hostility and sarcasm toward these people who have benefited you so much.

I don't know what world you live in, but I'm really, REALLY glad I'm not in it.
posted by Malor at 11:01 AM on January 19, 2011




« Older Cee Lo's 'Fuck You' in ASL   |   You were alive this morning, when the sun came up.... Newer »


This thread has been archived and is closed to new comments