Hacker News Reader

Comment by ZeroGravitas | Hacker News Reader

So remember there was a flurry of talk about what was going to kill Flash a few months back?

This (meaning Google's fairly clear intention to make some kind of royalty free video codec from what it's just bought) has just killed Flash and moved the web forward about 5 years.

I wonder if they'll create a new standard around the technology (perhaps pair it with the Ogg container format or with Vorbis audio, run it past some standards org like IETF) and/or release it as is.

It's a fairly big move for Linux Multimedia support too.

View on HN

I don't think this is going to kill flash, or at least not for a while. The speculation that Google is going to switch YouTube over to HTML5 any time soon is quite far fetched. They would be alienating a large part of their user base.

What this does provide is the possibility to have a universally supported HTML5 video codec. At the moment, there isn't a single one that is going to be supported by all browsers. Without a well supported codec, HTML5 video tags are basically useless in production. This new deal doesn't automatically kill flash, but it does somewhat level the playing field between flash and HTML5 in terms of video.

Well, considering bandwidth is more important to google than storage, they could take the more obvious approach and just choose the appropriate method and codec for a person's web browser.

Google is losing the Youtube money on streaming bandwidth. 1+1=2

Youtube already has specialist interfaces for smartphones, iPhones, Big screen TVs, and 6 or 7 different Flash versions based on quality/bitrate, size and codec. They can easily add another using HTML5, particularly if it provides a better experience for those using Google's Chrome browser or OS.

YouTube native on the iPhone was already a major blow to Flash (as was the whole touchscreen phone/computer paradigm). This will kill it (in conjunction with a bunch of other stuff Google is pushing under the umbrella of HTML5).

Video in Flash has always been an aberration created by dodgy patents, licensing considerations and competitive posturing of big companies, now the market can correct itself and Flash (for video and otherwise) is going to have to compete on its own merits. In brief: dead man walking.

I really hope you're right. I would love nothing more than to see flash go the way of .rm

One of On2's codecs, VP6, is already supported out of the box by Flash, potentially allowing them to provide the same video encoding to both HTML5 and Flash visitors.

It's not only about royalty's. For example, Apple happily pays the H264 fees for its iPhones and iPods. Having a free alternative won't make that stop, the reason they use H264 is because there are a lot of hardware acceleration chips out there which they can use.

As long as those aren't available for this new codec (or Ogg for that matter), and they aren't in most of the iPhones / iPods currently out there, Apple won't support this codec.

a) the whole hardware acceleration thing is 99% myth. The core elements of standard video codecs are all accelerated the same way. Apple (ironically) buying a combined hardware/software solution for its devices that it doesn't have the right or the capability to upgrade is a radically different proposition from no acceleration possible/available.

b) Apple was making those statements from a position of strength, they didn't have to be particularly convincing or true to work in that political forum. Though they're still holding a lot of cards, they're much weaker in a world with a technically acceptable, Google-backed, royalty-free codec in the wild. Previously it wasn't in Apple's interests to bother, now things might be different. It's not like Apple owned or controlled H.264 or any other Quicktime codec either.

c) one of the On2 sub-companies that Google just bought (called Hantro) builds video acceleration chips for mobile phones.

I'm not sure (a) is true. I've worked with advanced DSPs, like TI's DM-642. It's popular to do H.264 compression. The hardware acceleration in DSPs like this have nothing do to with H.264 codecs, you're right.

But the software libraries to do the best in class compression are H.264. And I'm not going to write my own compression algorithm. So the choice is lower quality or licensing libraries to take care of the compression.

The software for this DSP would certainly be a hell of a lot cheaper with a Google backed solution. And it might even be free.

Reply on news.ycombinator.com