Google and H.264 - Far From Hypocritical(blogs.computerworlduk.com)
blogs.computerworlduk.com
Google and H.264 - Far From Hypocritical
http://blogs.computerworlduk.com/simon-says/2011/01/google-and-h264---far-from-hypocritical/index.htm
7 comments
H.264 is already a part of every content producers workflow. It has very mature encoder support and it supported by every encoder that matters.
WebM's encoders that are available are very slow, and not supported or has very flaky support of most apps people are using out there.
What's going to happen, is chrome users are going to be served H.264 in a flash wrapper.
This move is not going to help kill flash, it's going to help extend it's life. If you think about it, it is not in googles best interest as an advertising company to kill flash. They'll have more trouble with content protection and are unable to overlay ads in fullscreen mode.
WebM's encoders that are available are very slow, and not supported or has very flaky support of most apps people are using out there.
What's going to happen, is chrome users are going to be served H.264 in a flash wrapper.
This move is not going to help kill flash, it's going to help extend it's life. If you think about it, it is not in googles best interest as an advertising company to kill flash. They'll have more trouble with content protection and are unable to overlay ads in fullscreen mode.
Actually, Apple does deserve some credit. H.264 is a standard in the video industry, along with Final Cut Pro (which revolutionized digital video production).
Flash adopted H.264 later.
Flash adopted H.264 later.
when in reality the major, basically only, reason it's used on the web is because of Adobe Flash
Could you elaborate on this or provide a citation? I thought Flash video was almost exclusively a proprietary variant of the H.263 (yes, that's not a typo) standard until the most recent release of Flash.
Could you elaborate on this or provide a citation? I thought Flash video was almost exclusively a proprietary variant of the H.263 (yes, that's not a typo) standard until the most recent release of Flash.
Originally "flash video" was Sorenson Spark/H.263, in 2005 they added On2's VP6 which was their high quality option until they added H.264 a couple of years later. Each step improved quality but mostly by adding decode complexity so as computers have got faster people have migrated to the heavier-weight codecs.
I would imagine that today all but legacy content served via Flash on the web is a standard H.264 file that can be easily repurposed for serving via HTML5 to Safari and newer iDevices (older ones only supported the lower profiles).
I would imagine that today all but legacy content served via Flash on the web is a standard H.264 file that can be easily repurposed for serving via HTML5 to Safari and newer iDevices (older ones only supported the lower profiles).
The other thing to remember about Flash though, is that they achieve something like 97% penetration very quickly. There are very few devices sitting out there w/ flash9 or earlier.
See this: http://www.adobe.com/products/player_census/flashplayer/vers...
See this: http://www.adobe.com/products/player_census/flashplayer/vers...
Exactly, which raises the question of why Adobe would bother transitioning a market that they completely dominated from one codec to another if they weren't feeling the pressure of H.264 from the mobile space (ZG's assertion appears to be that mobile in general — or, at least, iOS in particular — exerted no such pressure).
I think Adobe rarely gets accused of having their finger on the pulse of the mobile market around the year 2007. Instead H.264 was just an obvious choice, they would probably have gone with it in 2005 if the licensing situation was better, they certainly got abuse from some quarters for not doing so at that time.
My contention was that serving H.264 became feasible the day Flash started supporting it. The day before VP6 was a better choice because everyone had it installed and could therefore actually watch your video. The same is true now of H.264, it may be a wonderful codec and used in Blu-Ray and blah, blah, blah but the reason you can put it on a website and have 98% of people watch it is because it's supported by Flash. Sure Apple can force you to install Quicktime to watch their keynote, Microsoft can do the same with Silverlight, everyone else uses what's already out there on basically every machine, which is, and has been for a while, Flash.
My contention was that serving H.264 became feasible the day Flash started supporting it. The day before VP6 was a better choice because everyone had it installed and could therefore actually watch your video. The same is true now of H.264, it may be a wonderful codec and used in Blu-Ray and blah, blah, blah but the reason you can put it on a website and have 98% of people watch it is because it's supported by Flash. Sure Apple can force you to install Quicktime to watch their keynote, Microsoft can do the same with Silverlight, everyone else uses what's already out there on basically every machine, which is, and has been for a while, Flash.
Adobe makes their money on tools, not Flash. If there's going to be a new market, they'll jump on it. They said so many times. Currently they're working on a wysiwyg HTML5 editor, Flash to HTML5 converter, etc.
Another telling point (and an easy way to identify Apple fanbois) is that this isn't a complaint about proprietary plugin usage. This is a complaint about Flash. And that's where the sour grapes are. Anything that fanbois want to say about the inclusion of Flash in Chrome can be said about Java Applets, but there hasn't been the merest hint of a peep about Google's relationship with Oracle.
To be perfectly honest with you, your comment reads far better without the use of the word fanboi. That's name calling, and it sits at the lowest rung of Paul Graham's "how to disagree" scale.
Duly noted for future use.
I however do intend some derision for what i view as a piece of fairly serious intellectual dishonesty.
Not sure whether it'd be better to leave the comment up intact, given that you've commented at this point, or edit the term out.
I however do intend some derision for what i view as a piece of fairly serious intellectual dishonesty.
Not sure whether it'd be better to leave the comment up intact, given that you've commented at this point, or edit the term out.
Leave it there. The rest of your point is valuable, and Hn readers ought to be (ha!) open-minded enough to:
a) Read your point and agree with it even if "fanboi" isn't their favourite word, or; b) Read your point and respectfully dissent even if "fanboi" is their favourite word.
This isn't a "tea party."
a) Read your point and agree with it even if "fanboi" isn't their favourite word, or; b) Read your point and respectfully dissent even if "fanboi" is their favourite word.
This isn't a "tea party."
Of course it's a business decision by Google. It would be negligent to their shareholders if it weren't a good-faith business decision for Google. It makes plausible strategic business sense for all the reasons others have outlined. (I think it's a bad move, but hey I'm not running the billion dollar company.)
Same with Apple. It is too simplistic to say "Apple dumped Flash". There were three great business reasons for Apple to not have Flash on iOS. (1) A decent speed/energy Flash didn't exist, and Apple had no leverage to get Adobe to make one, (2) Apple does not want to depend on any outsiders to advance their platform, (3) Apple wants to differentiate iOS, so they want to have software target advanced iOS features and be exclusive. Cross-platform tools like Flash defeat that.
Folks, these are the biggest tech players in the world. I don't think either one gives a crap about standards or the free software movement any farther than it advances their business model. We've been in a lucky stretch where both Apple and Google want to commodify web access, so they've support (mostly) open standards. Now they turn their guns on each other and will seek to commodify each other's core products (for Apple, hardware; for Google, advertising).
Same with Apple. It is too simplistic to say "Apple dumped Flash". There were three great business reasons for Apple to not have Flash on iOS. (1) A decent speed/energy Flash didn't exist, and Apple had no leverage to get Adobe to make one, (2) Apple does not want to depend on any outsiders to advance their platform, (3) Apple wants to differentiate iOS, so they want to have software target advanced iOS features and be exclusive. Cross-platform tools like Flash defeat that.
Folks, these are the biggest tech players in the world. I don't think either one gives a crap about standards or the free software movement any farther than it advances their business model. We've been in a lucky stretch where both Apple and Google want to commodify web access, so they've support (mostly) open standards. Now they turn their guns on each other and will seek to commodify each other's core products (for Apple, hardware; for Google, advertising).
Google fan boys can try and rationalize this decision all they want, but a few things are known.
1. Most people saw this announcement as 'business' decision by Google
2. This was not a decision driven by what is best for the average computer/online user
3. Yay, Flash video! :(
1. Most people saw this announcement as 'business' decision by Google
2. This was not a decision driven by what is best for the average computer/online user
3. Yay, Flash video! :(
1. that's totally projecting. Read what they wrote, not what you think they mean. But if you really want to play that game, same can be said for Apple's decision to drop flash.
2. And dropping flash was good for average users?
3. Regardless of how much we all hate flash, flash video wasn't dead to begin with, or even dying yet. Sorry to pierce the reality distortion field.
2. And dropping flash was good for average users?
3. Regardless of how much we all hate flash, flash video wasn't dead to begin with, or even dying yet. Sorry to pierce the reality distortion field.
4. Google promotes the mediocre over the exceptional. WebM encodes are and will remain lower quality vs. H.264
If Google gets their wish and consumer video cameras switch to WebM, we all end up with lower quality recordings. Thanks! :(
If Google gets their wish and consumer video cameras switch to WebM, we all end up with lower quality recordings. Thanks! :(
I know I'm taking to a wall for most of you, but anyone with an open mind and more than 3 years of experience knows:
The reason we moved to H.264 is because it's a standard in the video world. It unified online video really well - everyone uses it and we were all tired of the fragmentation.
Flash is also a standard, particularly for multimedia. You can't have 99.7% penetration and not be a standard.
With all due respect to web standards: a standard is something that everyone uses. I think this is the best and most ironic line in the article:
"HTML5 is an emerging standard and anyone dealing with it has to expect change constantly."
The reason we moved to H.264 is because it's a standard in the video world. It unified online video really well - everyone uses it and we were all tired of the fragmentation.
Flash is also a standard, particularly for multimedia. You can't have 99.7% penetration and not be a standard.
With all due respect to web standards: a standard is something that everyone uses. I think this is the best and most ironic line in the article:
"HTML5 is an emerging standard and anyone dealing with it has to expect change constantly."
With all due respect to web standards: a standard is something that everyone uses.
It depends upon your priorities. Others would say that a standard is something that is completely & unambiguously specified so that interoperable implementations can be made. The terms under which such implementations are created is yet another concern. We probably need 3 orthogonal axes to plot various standards: ubiquity, specification, licensing terms. A "defacto standard" would score high in ubiquity but might have no specification (let alone one of quality) and might have bad or undefined licensing terms. Other things may have excellent specifications but niche existance. Something like H.264 has a specification, high penetration, but licensing terms that require certain participants to pay. Something like WebM is shooting for very liberal licensing terms, they are currently a niche but are angling for penetration — although I'm not sure if they have/intend-to-have a specification upon which interoperable implementations can be based.
Edit: there are also those who seems to put the greatest importance upon who controls the specification. They wouldn't accept PostScript as a standard, for example, because it isn't controlled by a standards bureaucracy —despite the clarity of its specification. I personally believe that unless we can openly acknowledge all of these components of the concept of "standardization" we'll have a lot of heat & no light when we're discussing it.
My personal leanings: I value a complete & ambiguous spec above all else, and loathe de facto standards that thrive in absence of one. I don't feel standards bodies are a must-have, but acknowledge that they can be nice. I also acknowledge that standards bodies can give the illusion of community control when in fact they provide no such guarantee. I'd rather not have to pay to implement a spec, but in some problem domains (such as codecs) I can understand why some would want to be rewarded for their work.
It depends upon your priorities. Others would say that a standard is something that is completely & unambiguously specified so that interoperable implementations can be made. The terms under which such implementations are created is yet another concern. We probably need 3 orthogonal axes to plot various standards: ubiquity, specification, licensing terms. A "defacto standard" would score high in ubiquity but might have no specification (let alone one of quality) and might have bad or undefined licensing terms. Other things may have excellent specifications but niche existance. Something like H.264 has a specification, high penetration, but licensing terms that require certain participants to pay. Something like WebM is shooting for very liberal licensing terms, they are currently a niche but are angling for penetration — although I'm not sure if they have/intend-to-have a specification upon which interoperable implementations can be based.
Edit: there are also those who seems to put the greatest importance upon who controls the specification. They wouldn't accept PostScript as a standard, for example, because it isn't controlled by a standards bureaucracy —despite the clarity of its specification. I personally believe that unless we can openly acknowledge all of these components of the concept of "standardization" we'll have a lot of heat & no light when we're discussing it.
My personal leanings: I value a complete & ambiguous spec above all else, and loathe de facto standards that thrive in absence of one. I don't feel standards bodies are a must-have, but acknowledge that they can be nice. I also acknowledge that standards bodies can give the illusion of community control when in fact they provide no such guarantee. I'd rather not have to pay to implement a spec, but in some problem domains (such as codecs) I can understand why some would want to be rewarded for their work.
I say tomato, you say Nix v. Hedden...
=)
=)
This further complicates building video into web sites with a significant IE user base. I've already seen mushrooming complexity with fellow devs trying to get Ogg into the mix using <video> tags, wanting to "do the right thing", and still ending up with a Flash fallback for Windows users.
HTML5 as we currently find it isn't making this simpler at all.
HTML5 as we currently find it isn't making this simpler at all.
Flash and h264 are completely different and are not at all equivalent.
Google doesn't support Flash. Google supports BROWSER PLUGINS. And Flash is by far the most popular plugin to the point of being ubiquitous. Google CAN'T drop support for Flash without dropping support for plugins. Google could, symbolically, stop bundling the Flash plugin. But really they believe in an open plugin standard which allows for browser extensions.
By comparison h264 is more like a proprietary, out of spec superset of the HTML standard. Building h264 support directly into the browser is equivalent to supporting IE specific markup. Or maybe even ActiveX. It just exposes Google to patent suits.
The real problem is the W3C needs to recommend a codec for the HTML5 video tag. Until they do there is no chance for standards compliant implementations to exist because there is no standard. h264 is not a solution because it's not part of the standard. And it won't become part of the standard until it loses the patent encumberance.
Google doesn't support Flash. Google supports BROWSER PLUGINS. And Flash is by far the most popular plugin to the point of being ubiquitous. Google CAN'T drop support for Flash without dropping support for plugins. Google could, symbolically, stop bundling the Flash plugin. But really they believe in an open plugin standard which allows for browser extensions.
By comparison h264 is more like a proprietary, out of spec superset of the HTML standard. Building h264 support directly into the browser is equivalent to supporting IE specific markup. Or maybe even ActiveX. It just exposes Google to patent suits.
The real problem is the W3C needs to recommend a codec for the HTML5 video tag. Until they do there is no chance for standards compliant implementations to exist because there is no standard. h264 is not a solution because it's not part of the standard. And it won't become part of the standard until it loses the patent encumberance.
So globally, in a couple of months when Google turns off H.264 in Chrome the HTML5 support for H.264 will stand at 5% (yes, that's not a typo) including all mobile browsing i.e. iOS which contributes less than a percent. Even in the US where Apple is strongest it only reaches 10%. Meanwhile in Europe WebM in HTML5 after Firefox 4 is released will be up at 60% (50% worldwide, 40% US).
So with numbers like 60% vs 5% why isn't WebM going to totally wipe the floor with H.264? How can people be complaining about Google ruining things and causing them so much extra work by taking away their 10% of the measly 15% share that H.264 has now? Short answer: Flash.
Flash currently delivers H.264 to 98% of the web. They've promised but not yet released VP8 support and can rapidly get it to a good 85% of the web within a year of release. What Google does with Chrome is really irrelevant in the face of that. It's more about sending a message. A message that roughly says "H.264 in HTML5 cannot work without Flash as a fallback or shiv, so a strategy to promote WebM via pushing people to Flash in the short-term is no less (or more) ridiculous". Apple's position is only tenable because of the actions of their two arch-frenemies: Adobe and, to a much lesser extent but growing as IE9 comes online, Microsoft and either or both may just see this as a chance to stick it to Apple and steal some glory for themselves.