StarDestroyer.Net BBS

Introducing YouTube HTML5 Supported Videos

For those of you who don't know what HTML5 is, let me give you a history. For a long time now Flash has dominated the online video world. If someone wanted to display a video on their website, they almost always used Flash to do it. For an even longer time, Flash has sucked unholy balls. It was never designed to play video, requires an absurd amount of processing power to display a simple 100 kbps video, frequently crashes (which brings down the entire browser), and uses the h.263 format which is a terrible codec originally designed in 1996 for videoconferencing.

So when W3C, the organization responsible for standardizing the way web pages are written and displayed, decided to update the standard for HTML, (to HTML5) a lot of people encouraged them to include an way for web pages to display video that didn't require a 3rd party plugin. Their response was the <video> element, which is used the same way as the <img> element (or the bbcode [img] tag). It was a very simple way to allow someone to include a video on their web page and have it playback properly on someone's computer.

The problem was that they needed to decide what video formats would be supported, like how the [img] tag supports .jpg, .gif, and .png. Mozilla (makers of Firefox) supported the Ogg Theora format since it was free to use and not encumbered by patents or trademarks. Apple (makers of Safari) supported H.264 since it is a better format and can stream higher quality videos even on slower connections. Google (makers of Chrome) supports including both. Mozilla is opposed to H.264 because it would require them to pay money to the makers of H.264 every time someone downloads Firefox, and also would require anyone who wants to make a video and put it online to pay. Apple is opposed to Theora (supposedly) because they are worried that if someone reveals that they have a patent that covers Theora they could get sued for infringement. (Alternate theory: Apple is one of the makers of H.264 and they want to get payed every time someone downloads a web browser or makes an online video)

The W3C responded to this argument by doing the worst possible thing they could have: stripping out all format support requirements and declaring that web browsers could support whatever they wanted to!

They could not have made a worse decision if they tried. Even removing the <video> element altogether would have been a better decision. The result of this is that the web is being divided into two competing standards in a way that is reminiscent of the old Netscape vs IE wars of the 1990s. Nowhere is this more apparent then on Youtube. As it says in the link at the top of the page, Youtube (owned by Google) just announced support for viewing via the HTML5 <video> element. But the catch is that it's only available in H.264. Which means if you have Firefox you're SOL and can't use the new feature.

Internet Explorer doesn't support anything from HTML5 yet, and won't until IE9. What they decide to support could decide the format war. My prediction? A big fat Fuck You to everyone involved and exclusive support of the Windows Media format. If that happens, the lulz will be epic.

Some notes: Flash supports H.263, VP6 and H.264. It is inefficient at decoding but Flash 10.1 will finally implement hardware video acceleration.

HTML5 is not really a W3C project; it pretty much took the whole WHATWG lock-and-stock after the failure of the XHTML standard.

At present, H.264 is available royalty-free until 31 Dec 2016 if you stream free content.

I'd expect IE9 to support H.264; Windows 7 already supports it out-of-the-box.

phongn wrote:Some notes: Flash supports H.263, VP6 and H.264. It is inefficient at decoding but Flash 10.1 will finally implement hardware video acceleration.

H.264 support was only introduced at the beginning of 2008. Most video sharing sites still use H.263 for most videos.

HTML5 is not really a W3C project; it pretty much took the whole WHATWG lock-and-stock after the failure of the XHTML standard.

My mistake, I thought WHATWG was part of W3C but after looking it up I see it's not.

At present, H.264 is available royalty-free until 31 Dec 2016 if you stream free content.

Which only was announced after HTML5 clusterfucked itself as a result of the arguing. Mozilla and other browser makers would still have to pay to implement a decoder, and content makers have to pay if they want to stream premium content.

I can't find any information on how this affects ad-supported content.

I'd expect IE9 to support H.264; Windows 7 already supports it out-of-the-box.

Maybe. This is probably going to be the last best chance for Microsoft to use it's old "embrace, extend, extinguish" tactic to rule the internet. I suppose it will be a great test of whether or not they're actually committed to embracing community standards.

Dominus Atheos wrote:H.264 support was only introduced at the beginning of 2008. Most video sharing sites still use H.263 for most videos.

Yes, but you were misleading; Flash supports a variety of video codecs.

Which only was announced after HTML5 clusterfucked itself as a result of the arguing. Mozilla and other browser makers would still have to pay to implement a decoder, and content makers have to pay if they want to stream premium content.

Mozilla could delegate to the host operating system instead. There is no reason why they must implement it themselves, surely? OSX and Windows 7 have it and the Linux guys will probably just download x264 anyways.

I can't find any information on how this affects ad-supported content.

Probably you're going to need to look at the license agreement itself.

Maybe. This is probably going to be the last best chance for Microsoft to use it's old "embrace, extend, extinguish" tactic to rule the internet. I suppose it will be a great test of whether or not they're actually committed to embracing community standards.

Microsoft doesn't give a whit about "community standards" - they do care about actual standards these days. They might ship some content in VC-1, I suppose.

phongn wrote:

Which only was announced after HTML5 clusterfucked itself as a result of the arguing. Mozilla and other browser makers would still have to pay to implement a decoder, and content makers have to pay if they want to stream premium content.

Mozilla could delegate to the host operating system instead. There is no reason why they must implement it themselves, surely? OSX and Windows 7 have it and the Linux guys will probably just download x264 anyways.

Yes, but Mozilla and Opera are trying to expand into mobile OSs as well. It's really hard to delegate a codec to an ARM-based embedded OS.

Maybe. This is probably going to be the last best chance for Microsoft to use it's old "embrace, extend, extinguish" tactic to rule the internet. I suppose it will be a great test of whether or not they're actually committed to embracing community standards.

Microsoft doesn't give a whit about "community standards" - they do care about actual standards these days. They might ship some content in VC-1, I suppose.

Well the actual standard doesn't require any codecs. Just so long as IE9 recognizes the <video> element it's HTML5 compliant, even if it doesn't play any videos at all.

Meh,

I dont have flash and could still view youtube in firefox. I just use the youtube downloader plugin to save the video as mp4.

Just to note, Safari is also supported, DA.

Dominus Atheos wrote:Yes, but Mozilla and Opera are trying to expand into mobile OSs as well. It's really hard to delegate a codec to an ARM-based embedded OS.

Those systems tend to have hardware-accelerated video playback as well, don't they? The host OS should present some sort of interface to the hardware.

So this means I can watch Youtube videos in Chrome without my CPU fan running after 5 minutes or so? I only have integrated Intel graphics, and the computer is a 7 year old Pentium 4 powered system.

So this means Youtube will work within EVE finally?

Dominus Atheos wrote:

phongn wrote:

Which only was announced after HTML5 clusterfucked itself as a result of the arguing. Mozilla and other browser makers would still have to pay to implement a decoder, and content makers have to pay if they want to stream premium content.

Mozilla could delegate to the host operating system instead. There is no reason why they must implement it themselves, surely? OSX and Windows 7 have it and the Linux guys will probably just download x264 anyways.

Yes, but Mozilla and Opera are trying to expand into mobile OSs as well. It's really hard to delegate a codec to an ARM-based embedded OS.

As has been said, there should be some sort of framework around for enabling this - almost all mobiles that can actually play video content have hardware h264 acceleration. Android, Moblin and Maemo and other Linux-derived mobile OSs can probably use the desktop version's potential GStreamer support since they mostly use that for video anyway. I believe there is also some sort of framework on Symbian for it. WinMo is probably the most deficient in this regard.

Maybe. This is probably going to be the last best chance for Microsoft to use it's old "embrace, extend, extinguish" tactic to rule the internet. I suppose it will be a great test of whether or not they're actually committed to embracing community standards.

Microsoft doesn't give a whit about "community standards" - they do care about actual standards these days. They might ship some content in VC-1, I suppose.

Well the actual standard doesn't require any codecs. Just so long as IE9 recognizes the <video> element it's HTML5 compliant, even if it doesn't play any videos at all.

Considering their actions and increasing codec neutrality in Silverlight, I don't think that will happen. Although I don't exactly see them supporting Theora out of the box, h264 support is pretty much guaranteed once they start supporting <video>. In fact, I can easily see them leveraging Silverlight and integrating it into IE as the plugin that handles <video> - the capabilities are there. Silverlight has excellent, hardware-accelerated, video playback - probably the best playback performance-wise you can get on the net right now, and they already support canned player UIs through Expression Encoder - selecting one of the most "neutral" of those and using it as the player for h264 <video> is probably the cheapest and simplest way MS could support the tag. Silverlight also has a Javascript and DOM bridge so the programmatic requirements of <video> (that is, being able to control the player through custom html element controls) can be trivially implemented.

Finally, as a personal note, the day Flash dies in a fire is the day the web will be much better off. Getting stuttering on a 2GHz dual core processor on 480p video because suddenly Flash decided to shit itself and grind the HDD is simply not acceptable. Its sad that enormously better technologies (Silverlight and HTML5) are much further behind in adoption.

phongn wrote:

Dominus Atheos wrote:Yes, but Mozilla and Opera are trying to expand into mobile OSs as well. It's really hard to delegate a codec to an ARM-based embedded OS.

Those systems tend to have hardware-accelerated video playback as well, don't they? The host OS should present some sort of interface to the hardware.

I don't think most smartphones have a h264 decoder chip, although I'm not sure. Obviously the iPhone does, but I don't know about Blackberries or Windows Mobile phones. Anyway h264 decoder chips also cost licensing fees, so that's still a cost to support what's supposed to be an open standard.

Netko wrote:

Dominus Atheos wrote:

phongn wrote:Mozilla could delegate to the host operating system instead. There is no reason why they must implement it themselves, surely? OSX and Windows 7 have it and the Linux guys will probably just download x264 anyways.

Yes, but Mozilla and Opera are trying to expand into mobile OSs as well. It's really hard to delegate a codec to an ARM-based embedded OS.

As has been said, there should be some sort of framework around for enabling this - almost all mobiles that can actually play video content have hardware h264 acceleration. Android, Moblin and Maemo and other Linux-derived mobile OSs can probably use the desktop version's potential GStreamer support since they mostly use that for video anyway. I believe there is also some sort of framework on Symbian for it. WinMo is probably the most deficient in this regard.

Online video shouldn't just be restricted to mobiles designed to play stand-alone movies, it should be able to be played back on even the wimpiest cell phone or internet device. And regardless, someone is going to have to pay to support it, which flies in the face of the concept of what an open standard should be.

Well the actual standard doesn't require any codecs. Just so long as IE9 recognizes the <video> element it's HTML5 compliant, even if it doesn't play any videos at all.

Considering their actions and increasing codec neutrality in Silverlight, I don't think that will happen. Although I don't exactly see them supporting Theora out of the box, h264 support is pretty much guaranteed once they start supporting <video>. In fact, I can easily see them leveraging Silverlight and integrating it into IE as the plugin that handles <video> - the capabilities are there. Silverlight has excellent, hardware-accelerated, video playback - probably the best playback performance-wise you can get on the net right now, and they already support canned player UIs through Expression Encoder - selecting one of the most "neutral" of those and using it as the player for h264 <video> is probably the cheapest and simplest way MS could support the tag. Silverlight also has a Javascript and DOM bridge so the programmatic requirements of <video> (that is, being able to control the player through custom html element controls) can be trivially implemented.

I think that in order for Microsoft to call IE9 HTML5 compliant it has to support <video> out of the box with out requiring any plugins. And they can't include Silverlight out of the box without stirring an antitrust mess.

Dominus Atheos wrote:I don't think most smartphones have a h264 decoder chip, although I'm not sure. Obviously the iPhone does, but I don't know about Blackberries or Windows Mobile phones. Anyway h264 decoder chips also cost licensing fees, so that's still a cost to support what's supposed to be an open standard.

Most smartphones have H.264 support built-in these days. It's just the standard for video today - and all the cost is more or less sunk since Blu Ray and HD-DVD supported it. I don't know of anyone who's bothered to produce a decoder chip for something like Theora.

Dominus Atheos wrote:Online video shouldn't just be restricted to mobiles designed to play stand-alone movies, it should be able to be played back on even the wimpiest cell phone or internet device. And regardless, someone is going to have to pay to support it, which flies in the face of the concept of what an open standard should be.

The hell it should. There needs to be some lowest-common-denominator, unless you're just going to give up and use something like MPEG-1. Furthermore, since when have open standards meant free standards?

I think that in order for Microsoft to call IE9 HTML5 compliant it has to support <video> out of the box with out requiring any plugins. And they can't include Silverlight out of the box without stirring an antitrust mess.

Microsoft can just have IE do the "you need Silverlight to display this content, would you like to install?" popup. Or delegate to the existing media framework.

phongn wrote:

Dominus Atheos wrote:I don't think most smartphones have a h264 decoder chip, although I'm not sure. Obviously the iPhone does, but I don't know about Blackberries or Windows Mobile phones. Anyway h264 decoder chips also cost licensing fees, so that's still a cost to support what's supposed to be an open standard.

Most smartphones have H.264 support built-in these days. It's just the standard for video today - and all the cost is more or less sunk since Blu Ray and HD-DVD supported it. I don't know of anyone who's bothered to produce a decoder chip for something like Theora.

Which is great for phones and devices that had video playback in mind when they were built, but most cell phones and small devices don't, and will have to add one if they want to be able to claim being HTML5 compliant.

For example, the Blackberry Curve doesn't support it.

Dominus Atheos wrote:Online video shouldn't just be restricted to mobiles designed to play stand-alone movies, it should be able to be played back on even the wimpiest cell phone or internet device. And regardless, someone is going to have to pay to support it, which flies in the face of the concept of what an open standard should be.

The hell it should. There needs to be some lowest-common-denominator, unless you're just going to give up and use something like MPEG-1.

No, because mpeg-1 would also require licensing fees. Whatever WHATWG comes up with and W3C certifies needs to be free to implement.

Furthermore, since when have open standards meant free standards?

Pretty much forever. At least since OSI defined it as

Patents: All patents essential to implementation of the standard MUST:

* be licensed under royalty-free terms for unrestricted use

Definitely since the European Union defined it as

The following are the minimal characteristics that a specification and its attendant documents
must have in order to be considered an open standard:

- The standard is adopted and will be maintained by a not-for-profit organisation, and its ongoing development occurs on the basis of an open decision-making procedureavailable to all interested parties (consensusor majority decision etc.).

-The standard has been published and the standard specification document is available either freely or at a nominal charge. It must be permissible to all to copy, distribute and use it for no fee or at a nominal fee.

- The intellectual property - i.e. patents possibly present - of (parts of) the standard is made irrevocably available on a royalty-free basis.

And although the W3C doesn't specifically mention Open Standards here, I still think this is relevant:

2. Licensing Goals for W3C Recommendations

In order to promote the widest adoption of Web standards, W3C seeks to issue Recommendations that can be implemented on a Royalty-Free (RF) basis. Subject to the conditions of this policy, W3C will not approve a Recommendation if it is aware that Essential Claims exist which are not available on Royalty-Free terms.

To this end, Working Group charters will include a reference to this policy and a requirement that specifications produced by the Working Group will be implementable on an RF basis, to the best ability of the Working Group and the Consortium.

I think that in order for Microsoft to call IE9 HTML5 compliant it has to support <video> out of the box with out requiring any plugins. And they can't include Silverlight out of the box without stirring an antitrust mess.

Microsoft can just have IE do the "you need Silverlight to display this content, would you like to install?" popup. Or delegate to the existing media framework.

Usually for a program or device to be compliant with a standard, it has to support it all by itself right out of the box. Otherwise Windows is already HTML5 compliant since you can install one of the other browsers that is fully HTML5 compliant.

Yeah, Microsoft has gone to such great pains to seperate IE from the fundementals of their OS, I can't see them easily making a step back to having Silverlight do all the work for the <video> tag, no matter how slight a step it might be, no matter how good Silverlight might be at doing all this stuff.

Its another anti-trust suit waiting to happen for that matter.

But I'm surprised the standard has become so 'open' in terms of video support and what it will allow and won't allow. This aint 1995 here...

And while Flash might be able to have different video formats, its still inside a container that is prone to burning down the whole damn website if you look at it wrong...

What I want to know is, who gives a flying fuck what Apple wants? Safari and the Mac have like what 5% of their respective markets? How can a company single company with a rather small presence in computing have such a great influence as to derail an entire standard?

Stargate Nerd wrote:What I want to know is, who gives a flying fuck what Apple wants? Safari and the Mac have like what 5% of their respective markets? How can a company single company with a rather small presence in computing have such a great influence as to derail an entire standard?

How much of the smart phone market does Apple represent?

I think the iPad will be the first REAL test of that - Apples anti-flash position. Most people just don't web browse on smart phones THAT much. The Pad though, if its being pushed as a direct competitor to netbooks and other devices that REALLY do internet browsing...

It will be interesting to see their response anyway

RedImperator wrote:

Stargate Nerd wrote:What I want to know is, who gives a flying fuck what Apple wants? Safari and the Mac have like what 5% of their respective markets? How can a company single company with a rather small presence in computing have such a great influence as to derail an entire standard?

How much of the smart phone market does Apple represent?

Quite a lot. But how big is mobile browsing when compared to desktop/laptop browsing anyway?

Chris OFarrell wrote:I think the iPad will be the first REAL test of that - Apples anti-flash position. Most people just don't web browse on smart phones THAT much. The Pad though, if its being pushed as a direct competitor to netbooks and other devices that REALLY do internet browsing...

It will be interesting to see their response anyway

That makes it even worse, not only does Apple shun Flash, they make the adoption of a competing format more difficult through their insistence on their OWN codec.

Tsk, tsk.

Stargate Nerd wrote:

RedImperator wrote:

Stargate Nerd wrote:What I want to know is, who gives a flying fuck what Apple wants? Safari and the Mac have like what 5% of their respective markets? How can a company single company with a rather small presence in computing have such a great influence as to derail an entire standard?

How much of the smart phone market does Apple represent?

Quite a lot. But how big is mobile browsing when compared to desktop/laptop browsing anyway?

Nowhere near, I would assume, but I would also assume it's the fastest growing segment. How close to saturated is the desktop/laptop market in the first world?

Dominus Atheos wrote:Which is great for phones and devices that had video playback in mind when they were built, but most cell phones and small devices don't, and will have to add one if they want to be able to claim being HTML5 compliant.

For example, the Blackberry Curve doesn't support it.

Look, we have to draw a line somewhere. You have to drop support at some point, that's just the way it goes.

Furthermore, you argue for an open codec (e.g. Theora) which isn't supported in hardware at all, anywhere! You can't at once argue "well, not all phones support H.264 in hardware" and then go "we should support Theora!"

Microsoft can just have IE do the "you need Silverlight to display this content, would you like to install?" popup. Or delegate to the existing media framework.

Usually for a program or device to be compliant with a standard, it has to support it all by itself right out of the box. Otherwise Windows is already HTML5 compliant since you can install one of the other browsers that is fully HTML5 compliant.

Delegation to Windows Media would probably count since it ships that way out-of-the-box for all points and purposes. Microsoft could also simply not consider themselves compliant and delegate to Silverlight.

Stargate Nerd wrote:What I want to know is, who gives a flying fuck what Apple wants? Safari and the Mac have like what 5% of their respective markets? How can a company single company with a rather small presence in computing have such a great influence as to derail an entire standard?

It's not just Apple, you know. Google also prefers H.264 and they command the majority share of search (as well as own the major video destination). Apple's WebKit engine also powers the most popular mobile web browsers (Mobile Safari, Android Browser) and two popular desktop browsers (Safari, Chrome).

Stargate Nerd wrote:Quite a lot. But how big is mobile browsing when compared to desktop/laptop browsing anyway?

Small, but getting larger and larger ever since WebKit hit the mobile world.

That makes it even worse, not only does Apple shun Flash, they make the adoption of a competing format more difficult through their insistence on their OWN codec.

Tsk, tsk.

Apple does not own H.264.

RedImperator wrote:How close to saturated is the desktop/laptop market in the first world?

The desktop market is saturated and probably in decline; laptop sales keep growing (though with a greater shift towards low-cost commodity netbooks and such).

phongn wrote: It's not just Apple, you know. Google also prefers H.264 and they command the majority share of search (as well as own the major video destination). Apple's WebKit engine also powers the most popular mobile web browsers (Mobile Safari, Android Browser) and two popular desktop browsers (Safari, Chrome).

The desktop market is saturated and probably in decline; laptop sales keep growing (though with a greater shift towards low-cost commodity netbooks and such).

What is your mobile web browser market share estimate based on? Web statistics or phone market share? Since last time I checked Symbian still had more market share than iPhone, RIM and Android combined and according to latest estimates Nokia (Symbian + Maemo MeeGo) actually increased smart phone market share, although it is not clear whether it it was taken primarly from Apple, RIM or Windows Mobile Phone (it probably was not Android). If I had to guess I would say RIM and Windows.

The ultimate roadblock for mobile browsing is screen size. Although you certainly can use devices like the iPhone to surf, it's still not really the same as using a real computer screen of at least 12.1" diagonal. Solutions like the iPad are not significantly more mobile than netbooks and they suck at actually writing anything, so I highly doubt their ultimate viability, even if Apple's marketing prowess and inertia trasferred from iPhone and iPod touch will probably make the iPad at least an initial success.

Marcus Aurelius wrote:What is your mobile web browser market share estimate based on? Web statistics or phone market share? Since last time I checked Symbian still had more market share than iPhone, RIM and Android combined and according to latest estimates Nokia (Symbian + Maemo MeeGo) actually increased smart phone market share, although it is not clear whether it it was taken primarly from Apple, RIM or Windows Mobile Phone (it probably was not Android). If I had to guess I would say RIM and Windows.

Symbian uses WebKit as well and RIM's next browser will also be WebKit. Also, I was being sort of North America-centric, where Symbian has not penetrated.

Uh, sayng iphones suck for browsing and ipads suck for text entry makes people look pretty dumb. Amusingly the iPhone penetration leads to all kinds of sites having specific iPhone versions (or at least generic mobile viewports), which is good for everyone. I prefer doing banking on my phone because the website is better, and text entry on an ipad should be way better than iPhone.