Hacker News Reader

Comment by commandar | Hacker News Reader

I was discussing this with some people yesterday. I find it incredibly frustrating that the Grubers of the world have spent months attacking Google for supporting the evil, closed Flash, but then manage to spin Google dropping a closed format in favor of an open one as evil and using it as a roundabout way of continuing to hate on Flash.

The fact of the matter is that Flash is an entrenched, defacto standard and isn't going away any time soon. HTML5 isn't anywhere close to completely replacing Flash even if it were to disappear in a puff of logic right this moment.

HTML5 <video> is in its infancy, and isn't being perpetuated by sheer momentum like Flash. Further, H.264 was a complete non-starter for Mozilla, and Firefox holds nearly 20% of the market. Using H.264 for HTML5 <video> would have guaranteed market segmentation and hurt the chances of a truly open future.

This argument about dropping H.264 propagating Flash in the short term is just insanity to me. Flash is already here for the short term. We need to focus on our long-term options for moving to something more open, and the whining about this decision strikes me as totally myopic at best and blind fanboyism at worst.

View on HN

Your argument turns on Gruber hating Flash for being closed. He doesn't: He hates Flash because he thinks it's crap. Gruber is not an open source advocate. He has no objection to open standards, free software, or open source software, but they aren't his passion.

You're right that Flash is an entrenched, defacto standard. It is, however, much less of one than it was three years ago, and that's pretty much entirely due to one factor: iOS doesn't support it. The HTML <video> element has traction entirely due to the fact that it's the only way to play video from a web page on an iOS device. You don't have a choice between <video> or Flash: It's <video> or nothing.

Chrome dropping H.264 support isn't going to push adoption of the <video> element, because Flash continues to be a perfectly valid alternative method of providing video to Chrome. If Chrome dropped H.264 and Flash, that'd be a different story.

Gruber's position is entirely consistent: He supports moves that reduce the usage of Flash, and doesn't support ones that increase it. Google's choice to drop H.264 in Chrome will, in the short term at least, result in the latter. Acknowledging this fact is neither myopic nor fanboyism.

>You're right that Flash is an entrenched, defacto standard. It is, however, much less of one than it was three years ago, and that's pretty much entirely due to one factor: iOS doesn't support it. The HTML <video> element has traction entirely due to the fact that it's the only way to play video from a web page on an iOS device. You don't have a choice between <video> or Flash: It's <video> or nothing.

So is this debate about what's best for the future of open standards on the web, or is it about what's best for iOS?

>short term >myopic

S: (adj) short, shortsighted, unforesightful, myopic (lacking foresight or scope) "a short view of the problem"; "shortsighted policies"; "shortsighted critics derided the plan"; "myopic thinking"

http://wordnetweb.princeton.edu/perl/webwn?s=myopic

Sounds like exactly what I meant. It's focusing on short term consequences at the expense of long-term benefit.

As I said, dropping H.264 may result in the already entrenched Flash being used more in the near-term; in the long term it could help HTML5 <video> see better overall adoption.

help HTML5 <video> see better overall adoption

Who says it needs help?

> So is this debate about what's best for the future of open > standards on the web, or is it about what's best for iOS?

Dropping support for h.264 means slower adoption of <video> How's that better for the open standards? The most infuriating part of this is, that FSF and Google are so busy with their own agenda, that they forget users.

>Dropping support for h.264 means slower adoption of <video> How's that better for the open standards?

20% of the browser market was never going to adopt H.264. That's a fifth of the market cut out from using <video> entirely.

How in the world is that possibly good for web standards?

>The most infuriating part of this is, that FSF and Google are so busy with their own agenda, that they forget users.

That majority of the market has the entrenched Flash until HTML5 <video> gets sorted out as a standard everyone can use.

20% of the browser market was never going to adopt H.264. That's a fifth of the market cut out from using <video> entirely.

20% and dropping... Should we also stop the <video> train in its tracks because IE6 does not support it?

Dropping?

All lists for browser share that I have seen show FF as having either plateaued or started to drop by small percentages. If you add mobile browsers into the mix then even the plateau becomes a small drop.

And you really think that going to continue once Firefox 4 and Firefox for Android are released? Firefox is going to be relevant for a long while, so comparing it to IE6 is pretty disingenuous.

So is this debate about what's best for the future of open standards on the web, or is it about what's best for iOS?

Neither.

The debate is about what's best for the web, period. Some people consider open standards to be a priori superior. Others (like Gruber) do not. (I'm taking your usage of "open standard" to include freedom from patent encumbrance.)

Your last statement is a critical point in this debate: You assert that Chrome dropping H.264 support "could help HTML5 <video> see better adoption." This is not an uncontroversial statement, to put it mildly.

I frankly don't see how this move will aid adoption of the <video> element. Yesterday, sites could encode to H.264 and provide a <video> element to some browsers and a Flash container to others. Today, the same is true. Yesterday, sites could encode to both H.264 and WebM. Today, the same is true. The only thing that has changed is that the number of browsers that require a Flash wrapper to play H.264 content has increased.

Sites will potentially use the <video> element over a Flash container for several reasons:

* If they cannot use Flash. This is the case on iOS devices. iOS has done more to drive adoption of <video> than any other factor to date.

* For political reasons. Dropping H.264 support from Chrome will not change this; a site that avoids Flash because it is not free was presumably already encoding to WebM.

* Because the <video> element provides a better end-user experience.

For sites where this third case applies, dropping H.264 support may result in increased usage of WebM. However, note that this is not the same as increasing adoption of the <video> element--dropping H.264 may influence the codec choices of sites that have already chosen <video> over Flash, but it has no influence on sites that are willing to use Flash.

I would be interested to hear your arguments on how dropping H.264 could increase usage of the <video> element in the long term. I just don't see a case for it.

I do see a case that dropping H.264 might increase adoption of WebM. I am, however, dubious. So long as H.264 in a Flash wrapper is a viable fallback, I don't see any site choosing to encode to WebM as a result of this change. (Sites may choose to encode to WebM for other reasons--preference for freer standards or freedom from licensing fees--but in these cases, they would encode to WebM even if H.264 was available in Chrome.)

So, to summarize: You are trying to frame the debate in terms of freedom alone. It is entirely valid to prefer freer technologies and to make decisions based on freedom. This is not, however, the only possible debate, and it is not the debate John Gruber is engaging in. To accuse him of hypocrisy in supporting H.264 but not Flash is to misunderstand his argument.

>I would be interested to hear your arguments on how dropping H.264 could increase usage of the <video> element in the long term. I just don't see a case for it.

It's very simple - a significant share of the browser market was never going to support <video> with H.264. In which case you end up with a fragmented market and everybody's using wrappers like Flash anyhow.

Of the available formats, WebM is the one most likely to eventually get everyone on the same page. Without that, <video> never really becomes a standard.

It's been posted elsewhere in the thread several times, but if you have a few minutes, this is probably the single best article I've seen on the whole mess:

http://antimatter15.com/wp/2011/01/the-ambiguity-of-open-and...

If MS jumps on board, which wouldn't shock me at all, you've got 90% of the browser market signed on.

That doesn't explain how dropping support for H.264 drives adoption of <video>, however. Why would someone choose to use the <video> element because Chrome dropped support for H.264?

What would aid adoption of <video> is if there was a single codec that all browsers supported. That requires that browsers add codec support, however, not remove it--either Firefox and Opera would need to add H.264 support (unlikely), or the various platforms that don't support WebM would need to add it (impossible in some cases, since the hardware support isn't there).

>Why would someone choose to use the <video> element because Chrome dropped support for H.264?

Because given a sufficient timescale, WebM support in <video> might become ubiquitous. I've made it very clear I'm talking long-term, not tomorrow.

>What would aid adoption of <video> is if there was a single codec that all browsers supported.

There is a chance for this to eventually happen with WebM. It was never going to happen on any timescale with H.264. My point is simply playing on probabilities.

> (impossible in some cases, since the hardware support isn't there).

People keep harping on this, and while it may be unfortunate that some devices get left behind, as I've posted elsewhere in the thread, mobile currently only represents 3% of web traffic. The mobile web is still very much in its infancy and there's still plenty of time for things to change, even fundamentally in some places.

Because given a sufficient timescale, WebM support in <video> might become ubiquitous. I've made it very clear I'm talking long-term, not tomorrow.

I feel like we're talking in circles here. How does dropping support for H.264 make WebM support in other browsers more likely to become ubiquitous? How would retaining H.264 support make WebM support in other browsers less likely to be so?

H.264 for <video> guaranteed that there would never be a single, ubiquitous <video> format. The end result is that the best case you could hope for is some combination of H.264 and Flash wrapped video.

WebM doesn't have the same problems that were keeping that last 20% from adopting H.264. With Google, Mozilla, and Opera, you've got about 30% of the market on board with WebM already. If Microsoft can be convinced to adopt WebM (and, honestly, I don't think that's too much of a stretch), that's around 90% of the market signed up.

If that happens, your only holdout is Apple, and I think even Apple is going to have a hard time stonewalling the change given the way they've spent the last year preaching the wonders of the open web.

So, basically, what I'm getting at is that with WebM, there's at least a chance everyone could end up settling on a ubiquitous codec. With H.264, we already know it would never happen. An unknown probability greater than zero is a better prospect than a p=0, in my book.

>How would retaining H.264 support make WebM support in other browsers less likely to be so?

Fragmentation. <video> isn't a useful standard unless you know it's going to work. Some browsers supporting one codec and others another makes that impossible.

But, again, everything you say hinges on browsers (namely, IE and Safari) adding support for WebM, not Chrome dropping support for H.264.

If everyone supports WebM, then WebM wrapped in a <video> element is a viable choice.

Not everyone supports WebM at the moment.

How does Chrome dropping support for H.264 bring us closer to everyone supporting WebM? Support for the two codecs isn't mutually exclusive. Google is no more on board with WebM today than they were yesterday--they've removed support for H.264, not added it for WebM.

> Google is no more on board with WebM today than they were yesterday--they've removed support for H.264, not added it for WebM.

They've made it very clear the intention is to add it. I've said it's not happening overnight, and I really don't understand this attitude that if it hasn't happened instantly then everything is worse.

I'm replying here, since the threading has gone deep enough on your last response that I can't reply to it. Normally, I'd give up at this point, but I finally see where the misunderstanding is so I'd like to make one last comment. You said:

They've made it very clear the intention is to add [WebM]. I've said it's not happening overnight, and I really don't understand this attitude that if it hasn't happened instantly then everything is worse.

Chrome supports WebM right now.

Support for WebM and H.264 is not mutually exclusive. You can support one, you can support the other, or you can support both. That's what Chrome has done up until now: Support both. You could use either one, and it would play them back perfectly well.

Chrome is not adding support for WebM. They aren't adding support for it, because they already support it and have done so for as long as it existed.

Switching to a world in which everyone supports WebM in <video> tags does not require that a single browser drop support for any format--not H.264, not Theora, not anything.

Actually, Gruber hates Flash because Apple hates Flash. He had no problem with Flash until Apple decided to ban it from its ecosystem. Gruber's love for all things Apple is consistent and well documented.

The bottom line is that while a subpar technology, Flash is really not that crappy. It delivers rich content without any glitches to millions of desktops every day and while its implementation on phones is still not great, there is no technical reason why it won't improve to the same level we see on the desktops.

Here is why Flash is not going anywhere: the anger against it is not strong enough nor shared by enough users. Most web surfers are pretty happy about it, and as long as that doesn't change, Flash won't die.

I'm guessing your experience with Flash on the desktop is on Windows? I've never had significant problems with it there. On the other hand, my experience is that on a Mac (or worse, Linux), Flash is slow, crash-prone, a major CPU and RAM hog, tends to steal keyboard and/or mouse focus in weird ways when you least expect it, and just generally does not provide a very nice user experience.

This has nothing to do with Apple "crippling" Flash or anything like that either; if anything, Adobe has been improving the Mac version of Flash since before Apple started publicly ragging on it. They've still got quite a ways to go though.

Your argument turns on Gruber hating Flash for being closed. He doesn't: He hates Flash because he thinks it's crap

Not entirely:

Let me be clear, though: there is nothing wrong with playing a video in Flash. I mean that seriously, no sarcasm. What there’s something wrong with is requiring Flash Player to play video. That’s the whole point of the HTML5 <video> element: to enable web video without requiring the use of proprietary plugins.

From http://daringfireball.net/2011/01/practical_vs_idealistic

> I find it incredibly frustrating that the Grubers of the world have spent months attacking Google for supporting the evil, closed Flash, but then manage to spin Google dropping a closed format in favor of an open one as evil and using it as a roundabout way of continuing to hate on Flash.

You just simply cannot deny the audacity of Gruber's comments about WebM/H.264/Flash. They are completely self-serving.

I think if Gruber allowed comments on his blog, or had to talk about WebM vs. H.264 in an engagement, he'd have trouble shoving the pro-H.264 side of this debate down people's throats.

While he occasionally makes some valid points, Gruber tends to pick and chose which articles to post that'll support his esoteric rhetoric.

I agree Google is being self serving, but I also believe in this case their interests align with the average internet consumer.

For sure. Google and Apple both play the good for humanity card and both have had some failure and success with aligning their strategies with the people.

My only wish is that we identify the difference between tribe-mentality and logical necessity.

cookiecaper said it best here: http://news.ycombinator.com/item?id=2105570

I generally like Gruber and think he tends to weigh his thoughts and arguments very well - but I still can't forgive him for buying MPEG LA's PR release[1] hook and sink[2]. Since when has it been a good idea to take a press release at face value[3]?

He hasn't recanted this complete acquiescence, so I am under the impression that he sincerely believes them - and only on the grounds of that single release.

Maybe he was biased, too cute for his own good, had a brain fart, or is too stubborn to accede. I really don't know, but this is one case where I can't really blame Gruber haters for framing his writing and philosophy in a certain way.

[1]: http://www.businesswire.com/news/home/20100825006629/en

[2]: http://daringfireball.net/linked/2010/08/26/h264

[3]: http://shaver.off.net/diary/2010/08/27/free-as-in-smokescree...

    (Fixed a typo: "thought" to "thoughts". Don't read anything into it!)

Gruber stays in my RSS feed because he can be incredibly insightful at times, but when he decides to go off on one of these insane lines of thought, he really goes for it.

The last one I remember was when he spent weeks snarking about how Google was somehow being dishonest about not using private APIs in Android. The reality was Gruber hadn't spent 5 minutes looking at the intro to the Android SDK document, and fundamentally misunderstood how Android is architected. What really made it ridiculous is that he kept on with it after having been corrected on the matter, with one of those corrections coming from no less than Tim Bray.

Basically, Gruber is great when he sticks to Apple analysis; when he strays into discussing the competition he tends to enthusiastically dig himself a hole.

EDIT:

Here's that response from Tim Bray I was referring to:

http://www.tbray.org/ongoing/When/201x/2010/05/06/Private-AP...

Gruber was conflating internal datastores with private APIs. The official Android SMS app uses public APIs and then stores messages in its own internal datastore. Bray was cautioning app developers that they shouldn't access that store because the SMS app is not part of the core framework, can be replaced by other applications, and is not guaranteed to be there on all devices as a result.

Gruber managed to turn that into constant snarking about Google lying about only using public APIs.

I wish I could upvote you a dozen times. This exactly expresses my own feelings about DF. Every time he goes off on one of these snarky tirades I resolve to unsubscribe in my RSS feeder. He usually then redeems himself with an insightful, technically clueful run.

The most irritating thing about the h264 rant is that there are some valid points in there: content producers would have to encode content twice, lack of hardware decoding, pushing video back to Flash, proprietary nature of VP8's development. But these all seem like short-term problems to me. Hardware decoding will happen and Flash will support WebM.

The biggest long-term sticking issue to me is Apple. They have a lot of weight to throw around with iPhones and iPads, and it seems h264 is unavoidable there. Google could throw its own weight around by switching all of YouTube to WebM, but that would take some time as hardware and software catch up. In the meanwhile, Apple is selling millions upon millions of h264-only iOS devices.

And there's the patent issue too. It'd be nice of Google indemnified WebM, but they also have enough patents that they could cause problems for MPEG-LA too.

Google supports Adobe Flash, and has dropped support for Google's implementation of H.264 inside Chrome. One is a plugin. One was not.

I agree with your comment, and I find the Flash/H.264/Chrome argument to be childish at best.

If an Apple or Microsoft blogger finds themselves taking a more hardline stance than the FSF, it's easy to conclude that they're trolling/astroturfing.

However if we're talking "childish", then the "Plugins! I thought the point of HTML5 was to avoid plugins" argument wins by a mile. People, including Google, have been pushing HTML5 via various plugins, fallbacks and shims for a good while (e.g. Chrome Frame, Google Gears, Flash Vorbis decode, Java Theora decode etc.)

> However if we're talking "childish", then the "Plugins! I thought the point of HTML5 was to avoid plugins" argument wins by a mile.

Which doesn't challenge anything I said. In fact, you can remove plugins from you browser, including Flash, right now.

Sorry, I was agreeing with you about Flash and then going off on a tangent without making it clear I was meaning a different "plugin" argument. I was referring to coverage such as:

So Much For Standards, Google Says WebM Plugins Coming Soon For Safari And IE9

http://techcrunch.com/2011/01/14/webm-plugins/

Now that makes sense. I was reading it completely differently. Which is also why I probably couldn't find anything wrong with what you were saying.

Sorry for the confusion.

After reading this: http://antimatter15.com/wp/2011/01/the-ambiguity-of-open-and... It sounds like the plugins with be in the OS and not the browser. So it will be like downloading a codec for your Mac or PC not like installing Flash on in your browser.

The big distinction here is that it will be native to the browser because IE and safari use whatever codecs are in the OS instead of providing their own.

The hatred for plugins is ridiculous in general. One of the best things about software is that it can be made freely extensible, and browsers are no exception.

This can be seen pretty easily on Linux builds of Chrome/Chromium where it links to libflashplayer.so. In fact, when Flash crashes on Linux, Chrome pops up with an infobar telling you that libflashplayer.so has crashed. :)

Everyone is debating here as if Google's decision to promote WebM was static. That's not the case. This is step 1 in a process that could rid the world of Flash.

We need: 1 - Google to publicly put out a strong message (has already happened) 2 - iOS to support WebM (yet to happen) 3 - IE, and others to support WebM. MS has pledged support in IE 9; Safari will follow iOS. Firefox will. 4 - YouTube to gradually switch content from Flash to WebM. Maybe HD content is only going to be available in WebM for starters. Google can do a gradual switch over a course of 5 years or something.

So: Yey, Google! Even big evil companies can get it right sometimes.

Note also that Google, as the owner of YouTube, can kind of strong-arm everyone into supporting WebM. And by everyone I mean Apple which has a traditionally high resistance to any technology it doesn't control (or at least has a say in).

HTML5 <video> may be in its infancy, but H264 is a huge established technology, and tossing it out will be painful for a lot of people.

The best possible outcome of this exercise (and to its credit perhaps this is Google's intention) is for MPEG-LA to make H264 royalty free in perpetuity, which would eliminate the need to switch to WebM. Otherwise we're forced to make a whole bunch of currently useful devices a whole lot less useful in the interest of one day making a (by then) obsolete codec into a universal standard.

"Further, H.264 was a complete non-starter for Mozilla, and Firefox holds nearly 20% of the market. Using H.264 for HTML5 <video> would have guaranteed market segmentation and hurt the chances of a truly open future"

Except Internet Explorer will continue to be the highest-marketshare browser for the foreseeable future, and IE9 won't ship with VP8 (though you can install a codec for it, I doubt regular people will). It does ship with H.264.

Maybe Microsoft can be convinced to ship VP8/WebM in IE.

>Maybe Microsoft can be convinced to ship VP8/WebM in IE.

I certainly wouldn't discount it. MS made the H.264 decision well before Google announced WebM. It wouldn't shock me to see them cut a deal with Google if for no other reason than to put Apple at a competitive disadvantage.

How would it put Apple at a competitive disadvantage? Safari uses Quicktime for video playback. If WebM becomes important, it will be easy for Mac users to get support for it.

If you extend last years trend for another 12 months then (according to Statcounters figures) IE, Firefox and Chrome will all have 30% of the market each, with IE heading down, Firefox level, and Chrome going up.

Which means, somewhat surprisingly, that IE could be 3rd globally in the relatively near future. In fact the IE drop/Chrome rise appears to be accelerating, so it could happen sooner.

http://gs.statcounter.com/#browser-ww-monthly-200812-201012

But that's all versions of IE, IE9 won't be available on XP, and historically it's taken years for new versions to penetrate, so it'll be a much weaker force than Chrome and Firefox which complete updates in a few weeks and months respectively.

Microsoft getting on board would be nice, but Adobe getting WebM into Flash would take care of the older versions of IE that are hanging around too.

it's really about software patents & about their iThings being unable to play WebM

What's Mozilla's market share on mobile devices?

I don't know, but mobile only represents about 3% of web traffic.

http://www.quirksmode.org/blog/archives/2011/01/2010_mobile_...

How do you expect this number to change?

With time. The point is that both HTML5 <video> and mobile web are in their infancy. Browser desktop share is now and it seems foolish to ignore 20% of the market in favor of a segment that accounts for 3% of traffic.

h.264 is also now.

Reply on news.ycombinator.com