While your expensive super-duper video authoring toolchain does not now support WebM, if Google has its way, it one day will and you'll encode WebM/VP8 just like you encode H.264 today.
The lack of scriptability seems to stem from using tools that are hard to script. ffmpeg is just fine for most people that are going to be using the <video> tag. If your tools are hard to script, you may want to ask the vendor for better tools.
You're right that support for WebM is not as wide as is support for H.264. The theory is that it soon will be.
Nobody is forcing you to convert your H.264 videos today. If you want to wait until support is more widespread, then that's just fine. Chrome is not some devastating loss for the <video> tag here, its userbase is relatively minor.
Also, do you have any merit behind your statement that VP8 is a step backward? I'm aware of Dark_Shikari's post on the matter. I don't think it definitively shows that VP8 < H.264. Do you have anything else? Some sources have claimed VP8 is much better.
You cast this as if investment and legacy are irrelevant quibbles, when in fact they're commercial considerations of the kind that are pretty relevant when you want to do something like establish the video tag.
> While your expensive super-duper video authoring toolchain does not now support WebM, if Google has its way, it one day will and you'll encode WebM/VP8 just like you encode H.264 today.
This is a big assumption, and right now the only reason you would do it so you can serve ... well, virtually nothing. One day soon, three browsers in HTML5. You're certainly not reaching anyone new. By contrast, h.264 is well beyond the tipping point, lets you reach all-new mobile audiences, and has a huge installed base, and a toolchain that runs from home-level to corporation.
The drive for HTML5 video depends entirely on making it fit as closely as possible to the path of least resistance that is Flash Player. A decision like this makes HTML5 more expensive and more complicated for people to support, which simply means they won't.
> Some sources have claimed VP8 is much better.
Cite.
If you read Dark Shikari's analysis carefully, he says that VP8 is better than H.264 Baseline Profile.
As it happens, H.264 Baseline Profile is all that the "hardware acceleration" in most mobile devices handles (see http://weblogs.mozillazine.org/roc/archives/2010/05/its_a_re... ).
So if you want to get better quality than VP8 out of H.264 on the web you need to either lose hardware acceleration or get different hardware (I'm not sure how much of this can be done with the existing hardware and different firmware, and how much actually needs new hardware; the parts of H.264 non-baseline that are not hardware accelerated are the more complicated ones, obviously).
Of course VP8 also requires changes to the decoding hardware/firmware; I also can't tell you whether those changes would be such that the result would be able to handle non-baseline H.264.
But the upshot is that current VP8 video on the web can easily be higher quality per bit than current H.264 video on the web, because the latter is self-constrained by hardware support and the former is not worrying about it for the moment, since it's not there anyway.
All of which is a far cry from "VP8 is much better". ;)
Many of your arguments are valid against WebM (basically lack of maturity, support) but the exact same arguments can be made about HTML5 video with any codec (not supported by the majority of bowsers, it's not a standard yet, where is the DRM, hardware acceleration is patchy, captioning support is still sketchy, easy switching between streams etc.)
The set of those who oppose Flash as a closed web technology yet don't believe in royalty-free web standards is much larger than I thought it would be. But it seems to be mostly the cultier end of the Mac fanbase. It's a bit of a shame that they can't just come out and say "this affects me because Apple supports neither the current defacto standard H.264 via Flash, nor the coming de jure standard WebM via HTML5 on my portable iOS devices". (Note that the more normal Apple users were already annoyed, at Apple, for the first half of that lack of support on their iPads.)
Flash is in no-way teetering on the brink, and this decision by Google has minimal impact compared with the fact that 99% of web video is served via Flash. The only thing likely to make a dent in that number is changes at Youtube. So if you really hate Flash that much, I'd stay on their good side.
It does that by ignoring the state of play of video content, which is what could have driven the change to HTML5. To those guys it's as though they removed support for JPEG, frankly.
What's more, there are real, sensible, pragmatic arguments against doing this, and they don't all boil down to "wah but I has iPad". If anything, the iOS users are going to be absolutely fine out of this, because they're already where people with video content need them to be.
The only losers are people who would like HTML5's video tag to replace Flash as widely as possible. That's why Adobe is delighted.
Of course you can plug in another transcoder on the end of the output chain, but why not just serve the h264 holdouts yet more Flash and keep your existing archive and workflow untouched?
This whole debate is stupid and of course WebM exclusivity is the wrong way to go.
An "of course" is a really lame way to enter a debate, because it says that everyone take the debating so far seriously are idiots, which I don't believe the rest of HN are. The other alternative is that it's much more nuanced than "of course" allows, you just haven't understood the debate.
> The only reason people are complaining is that their iPhones have H.264 hardware acceleration
No, people are complaining for many other valid reasons. Like: the digital video world runs on H.264, it has deep, complicated, expensive internal toolchains to support it, legacy archives encoded in it, basically entire businesses built around it. The video production world is far larger and more complex than you're picturing it. A "script" isn't nearly enough to do what you're talking about; suggesting you just need to point ffmpeg at a hard drive of stuff has dramatically underestimated the scale of what's involved.
Or, reasons like: they want the video tag to succeed, and this decision essentially entrenches Flash for another five years at least.
In an h.264 world the video tag had the combined support of three major desktop browsers, iOS, Android, a tonne of shipping mobile and embedded hardware, a Flash fallback and crucially it was an easy implementation for all the video content producers, with no re-encoding required. That's plenty enough to give it a good beachhead and momentum away from all Flash all the time.
In a WebM world, you have the support of Firefox, Opera and Chrome. No shipping mobile support. No shipping embedded support. No shipping Flash implementation. No video producer support. No video toolchain support. And it's a small technological step backwards, to boot!
> I find the pointless hipster-fanboy whining to be pretty grating, personally.
I find underinformed freetard passive-aggression to have no place here, personally.