The Practical vs. Idealistic Scenarios for the Near-Term Future of Online Video | Hacker News Reader

The Practical vs. Idealistic Scenarios for the Near-Term Future of Online Video

daringfireball.net

60 pointsdanilocampos40 comments

It's hard to disagree with his analysis of what the practical outcome will be.

Google's not stupid though, so there's either another piece of the puzzle still to be revealed or Google's aim with this is to throw a spanner in the works for Apple's various iOS projects. AirPlay comes to mind in particular as a major feature that would neutralised by reversing the trend towards standardising on h264.

This is ok. Google can do this if it wants. Just don't dress it up like it's some selfless contribution to the good of mankind.

So let me get this straight.

Apple supporting only h.264 and deliberately not supporting flash in order to force uptake is a good thing.

Google supporting only WebM and deliberately not supporting h.264 in order to force uptake is a bad thing.

This doesn't seem to be particularly logically compatible to me. Gruber seems to imply the first was ok because Apple got their first and it's already happened, but I'm sure that a lot of people with flash video delivery systems did not particularly enjoy converting to h.264, but they did it.

Considering Chrome's growing market share, plus Firefox and Opera, content providers /will/ be forced to offer a compatible video delivery system. If that has to be flash, so be it.

I agree. Gruber says:

"Supporting WebM and H.264 is better than supporting only one or the other, in my book. But if you’re only going to support one, I say support the one that is in wide use, with extensive wide-ranging support from camera makers, mobile playback devices, and online video services."

But I seem to remember that he wasn't that keen on Apple supporting Flash, even though it was in wide use by mobile playback devices, and online video services.

Those aren't criteria for judging the fitness all web technologies, they're criteria for deciding between H.264 and WebM, should one feel obliged to choose.

There's no hypocrisy here, because you can take it as a given that Gruber believes that Flash sucks, and not just because he's an iOS partisan. All Mac users were subject to years of abuse by the Flash plugin. I gratefully welcome its decline.

But I seem to remember that he wasn't that keen on Apple supporting Flash, even though it was in wide use by mobile playback devices, and online video services.

Hardware vendors like camera and phone manufacturers aren't substitutable for online video services. And prior to the release of Android 2.3, where was Flash in any measurable use by "mobile playback devices"?

> I'm sure that a lot of people with flash video delivery systems did not particularly enjoy converting to h.264

Flash is a wrapper, not a codec, and has in fact supported H.264 encoded video for some time. A "switch" from Flash to (likely MP4 wrapped) H.264 entails only removing the Flash wrapper, not a re-encode of the content. Apple not supporting Flash is not the same thing as Google not supporting H.264 (codec != container).

I know full well that Flash is not a codec; I never implied as such. The point is, many people served a non-h.264 codec through flash, and so they would not have just been able to supply the same source video through a <video> tag rather than through flash.

I don’t think Theora and VP8 were widely used before the first browsers gained support for the <video> tag. Does Chrome support other codecs than those two?

Flash...has in fact supported H.264 encoded video for some time.

Could someone offer a citation for this? My understanding has been that the codec typically used for Flash video has long been almost-H.263, not H.264. Wikipedia, for what it's worth, currently reflects that version of history, saying "Commonly, Flash Video FLV files contain video bit streams which are a proprietary variant of the H.263 video standard", and "The most recent public releases of Flash Player...also support H.264 video".

http://en.wikipedia.org/wiki/Flash_video

Does Wikipedia have history wrong here?

> Google supporting only WebM and deliberately not supporting h.264 in order to force uptake is a bad thing.

It's not that simple. It's bad because it cannot possibly have a beneficial effect, as Gruber demonstrated I think pretty well in his article.

Lets be a bit clear here, Apple's choice of supporting H.264 didn't force (many) people to change their encoding format. H.264 has long been the primary choice for video (outside of the Microsoft formats).

A lot of people are saying "boo Apple" for choosing H.264 for their iOS devices. This is despite the fact that it really the only viable choice for their users. VP8 was proprietary up until May of last year and Theora is really not comprable with respect to bandwidth and quality. There was not a viable open format until WebM and at this point H.264 is deeply entrenched.

Personally as long as the W3C doesn't declare a format for the video tag (like images) I'm okay with Google's move. I think its going to greatly slow adoption of the video tag however. At this point most content providers are going to stick with Flash online for their video players so they can then provide the same (unwrapped) video for their users on other devices. When video content providers are to enocode once in H.264 and use a flash player or encode twice to partially support a now crippled native video tag, I think they are going to stick with one codec, and use flash. Google's choice may strengthen WebM's position in the long term but its weakening native video in the short.

Gruber has said that he liked Google’s approach of including many codecs (WebM, Theora, h.264) in the recent Talk Show episode with Dan Benjamin. I’m consequently not sure whether he is supporting Apple’s decision to ship OS X Safari only with h.264, it doesn’t sound like he does. (He is certainly making less noise about it than about Google’s decision.)

This is probably his position on Safari for OS X, not Safari for iOS, since WebM hardware is not out yet (and it is consequently crazy to expect Apple to support WebM on mobile devices – Android phones also don’t).

Apple supporting only h.264 and deliberately not supporting flash in order to force uptake is a good thing.

Google supporting only WebM and deliberately not supporting h.264 in order to force uptake is a bad thing.

That's a very poor reading of Gruber's argument. Instead:

Apple supporting only h.264 and deliberately not supporting Flash in order to force uptake is a good thing.

Google supporting only WebM and deliberately not supporting H.264 will not force uptake of WebM, and will instead drive the continued use of Flash for playing video.

Perhaps, though the disconnect here is that open source fanboys pretty much dislike H.264 as much as flash. Neither is open source by OSI standards: http://www.opensource.org/docs/osd. All the screaming about running from flash seems kind of silly when you point to something like H.264 as the alternative.

It is a standard, but that means little with all of the royalty-bearing patents surrounding it. It is royalty-free in a limited use case, but still can't be included on your typical linux distribution legally.

Not that I think Google is being altruistic. They obviously have a vested interest in controlling the video format youtube uses. Now is the time to bring a new format to the table, too. Gives them some breathing space to fight it out before worrying about paying royalties for H.264 or its successor.

I'll happily deal with a couple more years of flash (and honestly, we were going to anyway, ie9 isn't going to sweep away 7 & 8 anytime soon) to get WebM as a standard instead of H.264.

I've tried to stay out of this debate so far but I'd like to add one point: In my experience, the performance of H.264 playback in Adobe Flash on low-end computers is simply unacceptable. Pretty much any other option offers way much better performance: HTML5 H.264 in Chrome, Safari or Silverlight H.264 playback. I don't know whether it's their version of the H.264 codec or something else in their rendering pipeline but it doesn't really matter anyway.

We're a business-to-business shop that need to deliver web-based video content to our business clients. What we care about is our clients reliably being able to play back video on any of the thousands of their shitty underpowered old Dells. This is why they pay us so we get to come to work tomorrow. Still, we like to think that we do the right thing if we can so we would love to use standard, open technologies but this whole mess around HTML5 video codecs (and the fact that Flash is just way too slow on our clients' machines) forces us to use Silverlight instead which unfortunately also means that we can't support Linux boxes or any portable devices.

The only way I can see Google exercising any leverage here is by using the nuclear option and requiring WebM to view YouTube (via <video> or a newer version of Flash). YouTube is enough of a draw that WebM support would become more or less universal, via Flash upgrades and (presumably) native support on Android.

At which point, Apple would find itself pretty lonely on the H264 ice-flow.

Of course, every single mobile device currently in existence would become collateral damage, since WebM hardware acceleration support doesn't exist yet. (And hey, who knows how successful, cheap, or easy it will end up being.)

Seems kind of crazy to me, though, just like most of the "we've got to restrict your freedom to save your freedom" arguments (e.g., the recent GPL takedown of VLC from the App Store).

But I'm a pragmatist and tend the believe the future plays out in shades of grey rather than black-and-white, and am to be suspicious of causing certain harm now to increase the odds of avoiding some Dystopian future (and it's not like patents last forever!).

They don't have to go nuclear, they just have to serve youtube content in higher resolutions in WebM than in H.264.

That's an interesting idea. I wonder if people care enough about video quality on YouTube(!) for it to have much of an effect, tho...

They would when YouTube adds paid for content.

I'm an end-user, not a content provider. I don't give a rat's ass how my video is served, so long as it's pretty and fast. I want my video served in ≥720p with no buffering, as YouTube does currently. To this end, H.264 and WebM both work fine. In fact, so does flash. The only problem I have with flash is my laptop gets really hot and my battery drains rapidly every time I use it for extended periods.

Now, as the average end-user, I don't care about which browser supports which format. I either use IE or Safari because that's what came with my system, or Chrome/Firefox because that's what my friend who's good with computers recommended. I'm not likely to switch browsers because of video issues unless one just plain doesn't work.

So what does Google's dropping H.264 mean for the consumer and, by extension, adoption? Not much. The consumer will have to suffer through hot laptops and short battery life, and most won't know why. As Gruber points out, this will not encourage providers to support WebM, rather they'll simply keep serving up flash.

I don't see a winning path, here. From a consumer standpoint, I'm perfectly happy with H.264. From a web standpoint, I wish we had been having this discussion 3 years ago. H.264 seems to have won.

The consumer will have to suffer through hot laptops and short battery life, and most won't know why.

I'll give you this for Safari, but I don't think IE9 will have enough of an adoption rate to make this a noticeable effect. Average people are unlikely to switch browsers, and they're also unlikely to upgrade (though I wish I had some statistics here). They're likely to be using IE6/7/8, without HMTL5 anyway, especially with IE9 not supporting XP.

From a web standpoint, I wish we had been having this discussion 3 years ago. H.264 seems to have won.

Wholeheartedly agreed.

Unless I'm missing something, IE9 will be shipping on new laptops, so it should have pretty wide usage.

Grubers apple loving fanboyism aside, he does make a valid point. But I think he is missing the strategy Google may have in this case.

1) Video sites are not going to re-encode all their video to WebM when the video still plays on Chrome (through flash). (grubers point)

2) Google will drop Flash video on YouTube if your browser supports WebM. Both Apple and Adobe are targeted by this move, obviously Apple because they are H.264 supporters and Adobe because.. well.. Flash.

3) As Flash becomes even less relevant (see previous point) Chrome will stop shipping with Adobe Flash, instead will require users to download the plugin like any other browser.

4) Major websites will start embracing HTML5, and with it, re-encode video to WebM, since it will work by default in most modern browsers (Firefox 4 supports WebM and Microsoft have stated they will support it in IE9 too).

I'm sure Google is thinking something along those lines. Apple is a smart company so they might have a trick or two themselves. There is nothing inherently wrong with Apple supporting WebM, except that Apple are probably sick of having Google products scattered throughout iOS (youtube, gmail and maps) and this would be another Google (VP8) product they would need to integrate.

Frankly, of all the stuff I've read about this, Gruber's stuff actually makes the most compelling points.

Now it seems popular amongst some here to dismiss Gruber as an Apple fanboy or simply to disagree with him simply because he likes Apple and they don't but that's a ludicrous position. The only question should be: do his questions have merit?

IMHO they clearly do.

The thing that absolutely floors me about this is that it prolongs the reliance on Flash, which as Gruber correctly points out, was one of the major factors behind the <video> tag in HTML5 (to get away from requiring proprietary plugins).

For the foreseeable future, the smart (and IMHO the only) thing for sites to do is support H.264. To not do so is to potentially alienate hundreds of millions of users, being those using the iDevices (not the least of which is the iPod Touch). Sites may choose to dual-encode in WebM but why would you? Double the processing time and double the storage, for what? A video you can play without using a Flash wrapper?

People keep pointing to the GIF precedent. That is tantalizingly similar but it suffers from the same kind of fallacies as comparisons of iOS vs Android to Apple vs Microsoft (Windows) do: it ignores the fact that the times they have a-changed.

Consider: I watch an awful lot of video on my iPad. Hardly any of it is in H.264 format. The solution? A $3-4 app called Air Video that transcodes to H.264 on the fly.

With the cheapness and abundance of computing power now, let alone 10+ years from now (which is when the tail end of the H.264 patents are and thus when the GIF-like profiteering scenario would presumably come to fruition), why do I care what format anything is in?

If MPEG-LA decided to increase the cost of an H.264 license by 10x for a site like Youtube I guarantee you that Google could re-encode every video on Youtube in somewhat reasonable time. And that's today.

H.264 is already (perpetually) free for the end user so the it's only the content providers who really matter in this equation.

Now I'm not arguing we should go completely in with H.264 and to hell with everything else. I'm all for providing users and sites with choice and for that reason, I fully support WebM.

What I don't support is removing the choice of H.264 from users or forcing them to keep using a closed and proprietary plugin (Flash obviously) to watch them.

The fact that Chrome bundles Flash is already almost enough reason for me to stop using it. Flash blocker extensions are suboptimal (eg tricks like Flash overlays on sites that kick off all the Flash). Completely uninstalling it is not an option (every page asks you "You are missing plugins, would you like to install them?").

For me this move may just mean I switch back to Firefox.

Lastly, I don't read this as an anti-Apple move at all. It's certainly political but (IMHO) not aimed at Apple specifically. It's more about the philosophy of an open Web. I fully support that. I simply question the wisdom of forcing subpar choices on end users to that end.

Also, some here like to point at Apple and ask "how is this different to Apple not supporting Flash?"

That one's easy: It was 3 years after the iPhone's release before Adobe had a full version of Flash that ran on ANY mobile platform (being Android ultimately). If you accept the assertion that Android Flash is sufficient (and that is debatable) why exactly is Apple being blamed for Adobe failing to deliver any kind of sufficient technical solution?

If you want to point the finger at anyone point it at Adobe. If they'd been serious they would've jailbroken various iDevices and demoed an iOS version of Flash years ago to make their point, had they had such a thing. The fact that either they didn't or they chose not to demonstrate it, whichever way it goes, means they dropped the ball.

Does Apple want Flash? Now? Clearly not and at this point it probably doesn't matter what Adobe does but years ago? Who knows? But criticizing someone for not shipping a nonexistent product is crazy.

you seem to be arguing that we should just keep using h264 unless things go bad, at which point we can always switch; do you not see the problem with that? even if MPEG-LA licensing fees never change, I'm with Mozilla (and google) on the fact that ensuring every video producer and provider will be paying license fees to put their content on the web for the next 17 years is suboptimal. codec choice is good, but until a royalty-free option is everywhere, I'd rather them play hardball.

speaking of mozilla, unless you run windows 7 and like microsoft firefox plugins, switching back to firefox won't do you much good on the h264 front.

You seem to be arguing that dropping H.264 <video> from Chrome will spur adoption of WebM. As many have pointed out, including the parent, a more likely scenario is it will spur adoption of Flash+H.264. Path of least resistance and all that.

Playing hardball is fine, but you've got to be sure you are actually playing and not merely whacking random spectators in the nuts.

I'm not sure this will happen. The only users who are unable to watch WebM are iPhone/iPad/Safari users. You can use WebM in IE9 (when you install the codec).

So as a webdeveloper the path of the least resistance for me will be WebM. You'll get support in Chrome, Firefox, Opera and IE9.

> H.264 is already (perpetually) free for the end user so the it's only the content providers who really matter in this equation.

definitely not true. h264 is free for use in free video streams. encoders, decoders, and commercial streaming all need licensing. browser (and OS) makers have just been paying for the encoder for you so far.

Googles's motivation is money. also control.

H.264 is not free. and youtube could be reclassified in 2016 to pay a further royalty increase than the maximum 10%. YouTube is the largest video service provider in the world.

On the other hand google owns WebM (spent 100M on it)

Also youtube is spending a pretty penny on flash "premium" codec support. Also not free, and could also potentially save money there pushing WebM.

it is really googles's only way to go

regarding Apple, a simple solution is to add webM support to IOS. they have some time to do it. will also save them money.

Moving to WebM may not save money for Apple. They gain money everytime someone pays MPEG LA for H.264 as they are one of the licensors. If H.264 loses widespread adoption Apple loses its share of the fees too.

I keep hearing this, but never with any figures.

Presuming Apple contributes a couple of patents to the pool, it doesn't seem likely that any income from MPEG LA would have any significance for them. Especially since they have to pay as a distributor of H.264 decoders anyway.

Apple is on both the list of licensors and the list of licensees. In other words, they own some of the patents in the pool, and they pay to use those patents in the pool that they do not own. In order for Apple to make a profit on patent licenses, they would have to either own a majority of patents in the pool (if all of the patents are of equal weight) or those patents that they do own would have to have greater weight than all of the patents that are owned by others.

It's hard to say whether Apple actually gains anything. For all we know, they might be happy to break even.

While I do not know the exact figures, there still is a good possibility for Apple to gain

They pay for H.264 license to distribute their decoder, but collect royalty for their patents from every other codec vendor, device maker and commercial video distributor in the world.

So, basically, promoting WebM is actually promoting Flash. Sure. Well put.

And then: so... what?

Am I the only one unsurprised that Google would fight Apple? (in convoluted ways sometimes)

The "what" being that, assuming the ideal is a move away from proprietary technologies like Flash, supporting WebM alone in the current environment is likely to have the opposite effect (and thus hurt web open standards, at least in the short to medium term).

I too liked Chrome's support of both WebM and H.264. Theirs is a disappointing move.

When the world gets into H.264 there's no way back but waiting for 20 years for patents to expire.

This was the last chance to at least try to keep the web video both open and free. It's a far shot, but it's at least a try. Before this move, the game seemed beyond lost, now it is only probably lost.

Flash will support WebM too, Skype is using it, so there are some chances. It's a very long way off, but now WebM has a fighting chance.

Too late. Seriously. The world got into H.264. We're here. It's it. What is WebM's fighting chance? Did you read the article? Do you have any counter points to Gruber's thesis? The web will continue to use H.264 and all Google is doing is ensuring that Chrome users (myself included) just get H.264 video in a flash wrapper. If we get WebM through Google/YouTube then great, but nobody else is going to bother serving it to us... why would they?

Only something incontestably better than h.264 will replace it. The consumer electronics industry doesn't waste resources going sideways, only forward.

As for your last statement: no. My bet is Flash gets WebM support natively next as part of a bargain between Adobe and Google. And that Android phones get WebM hardware decoders forced down upon them.

For Apple, they took a stance and now they risk Firewire/USB again.

adobe announced that they would add vp8 to flash months ago. as for a "bargain"...what would that even consist of? you can use and re-implement vp8 to your hearts content today without payng anyone.

Usually Google brings the big guns to their IO conference. Last year, the WebM project was announced. This year, some other surprise that will probably make this discussion irrelevant.

I'd rather spend time guessing on what the big surprise will be this year regarding web video than debating the effects of Chrome dropping h.264.

The ONLY thing I take from this move by Google is they have intention to take down h.264, and that in itself is enough to prepare myself for the possibility of serving 2 formats.

It's just one move in a series of moves Google has been and will be making.

Reply on news.ycombinator.com