Mozilla Plans H.264 Video for Desktop Firefox | Hacker News Reader

Mozilla Plans H.264 Video for Desktop Firefox

webmonkey.com

67 pointsglhaynes49 comments

H.264 is not a "proprietary video codec", it's a patent-encumbered codec. The spec is available for free on the ITU website(http://www.itu.int/rec/T-REC-H.264) and several open source implementations exist.

The spec is useless, even harmful, if 3 years from now MPEG LA decides to start charging content providers a licensing fee for their content, even if that content is distributed for free.

If we start using a codec for which the patents are owned by a company, and that company dictates the licensing terms for that codec, then aren't we handing control to how the content is distributed to said company?

what do we know about MPEG LA?

http://www.engadget.com/2010/08/26/mpeg-la-makes-h-264-video...

> as long as it's freely distributed

I was responding to the line "if 3 years from now MPEG LA decides to start charging content providers a licensing fee for their content, even if that content is distributed for free."

Aren't encoding and distribution fees both licensing fees faced by content providers? Your overall argument here isn't very strong.

If you are distributing free content, they have pledged not to charge you. It isn't so much an argument as a statement of fact.

They don't charge for the distribution but don't you have to create content to be a content provider?

Which is paid for by the folks making the software tools unless the OS paid the fee for the codec in which case you're good. The camera manufacture either paid the fee or shoots a format that can be converted on the PC.

That says absolutely nothing about h.264 decoder or encoder implementations, such as the decoders needed in web browsers to handle h.264 in <video> tags; that just says MPEG-LA won't charge people distributing h.264 video files.

The parent post didn't mention decoders / encoders, it mentioned free content. These articles[1][2] explain the costs and possible increases[3]. Apple, Microsoft, and Google have bought the license for their libraries on their platforms.

[1] http://www.zdnet.com/blog/bott/a-closer-look-at-the-costs-an...

[2] http://www.zdnet.com/blog/bott/h-264-patents-how-much-do-the...

[3] http://www.mpegla.com/main/programs/AVC/Pages/FAQ.aspx last question

Exactly. All they've said is "we promise we won't charge you for sending bits of h.264-encoded data across a wire". It's not clear that they legally could charge for that even if they wanted do!

And yet they've fooled vast numbers of people into thinking that h.264 is free-of-charge as a result. It's the most impressive piece of anti-FUD I've ever seen, sadly.

Proprietary, as in proprietor, as in owner of property, as in someone with property rights, as in government-granted monopolies, as in patents.

In fact in some contexts, proprietary is a synonym for patent e.g. patent medicine.

edit: from wiktionary, definition 3

Manufactured exclusively by the owner of intellectual property rights (IPR), as with a patent or trade secret.

> Unfortunately that dream has failed to pan out. Instead of proprietary plugins, the web ended up with proprietary video codecs, which has created a split in browser support for HTML5 video. Firefox and Opera support the open Ogg and WebM codecs, while Safari and Internet Explorer supported H.264.

Funny, it looks like a major browser was left out there (in fact the #1 browser in usage according to some metrics).

The article is pretty good overall, but is missing a key piece that is important to understand the background.

Yeah, from the "related" links at the bottom you can see this article from January 2011 entitled "Google Dropping H.264 Codec from Chrome Browser" ( http://www.webmonkey.com/2011/01/google-dropping-h-264-codec... ), which has still yet to come to pass.

Has anyone submitted a Chromium patch to remove H.264 support? I would be curious to see the code reviewer's feedback on the patch. :)

As far as I'm aware Chromium has never supported H.264.

I blame Google for this, too. They should've made Youtube play on WebM by default, with a fallback to Flash in browsers that don't support it. That would've forced the others to adopt WebM.

Forcing an inferior codec on users for political reasons is unreasonable.

Have you tried watching 720p WebM compared to 720p H264? The difference is jarring.

I watched the Amazing Spiderman trailer[1] in 720p in both formats - using youtube-dl - and I honestly couldn't tell the difference. The webM file is ~16% larger, though.

[1]: https://www.youtube.com/watch?v=atCfTRMyjGU

Well, Google even decided to bundle Flash into Chrome, so Flash is far from just being a "fallback" solution to them.

Does anyone have an opinion on whether Google's purchase of On2 makes any sense now with hindsight?

Did it affect the behaviour of the MPEGLA in any way? I know some of the tech has fed into WebRTC but that battle is far from over.

Was On2 worth the cash?

MPEG-LA? No. But MPEG itself is working on a royalty-free profile of AVC called WebVC: <http://mpeg.chiariglione.org/meetings/geneva11-1/geneva_pres.... That may not have happened without VP8.

Google acquired On2 in February of 2010. In August of 2010, the MPEG-LA "announced that H.264 encoded internet video that is free to end users will never be charged royalties." (http://en.wikipedia.org/wiki/H.264/MPEG-4_AVC#Patent_licensi...)

So... yeah, I'd say it profoundly affected the MPEG-LA's behavior. If Google didn't have an easy out with (what would become) WebM, their Youtube licensing would be much more expensive.

Whether that patent promise has any bearing on the Mozilla decision is a more complicated issue. People have strong opinions on both sides.

H.264 was already well entrenched when On2 was bought. Whether or not it was a good purchase will be seen more by a) how H.265 fares, b) how well AAC holds up against patent free mp3, Vorbis and Opus, c) how IETF's proposed Internet Video Codec and VP-Next fare (and they may potentially become a single project), and d) how many people are still paying for H.264 licences when it's no longer technically impressive, software-based companies in particular seem poised to jump off the upgrade treadmill early unlike earlier hardware based generations that could be milked for far longer.

This is great news. There will now be a de facto standard codec for HTML5 video!

Except it'll be the wrong one, sadly.

Exactly, this news is hardly any good, it's just convenient.

For the vast majority of people with any stake in video on the web, convenient is good.

I hope content publishers will find publishing HTML5 video convenient enough to drop Flash. Once publishers are all on board the HTML5 video train, they have the opportunity to experiment with new (and unencumbered) codecs without affecting their user experience.

Lots of video will be played on mobile devices, so any codecs without hardware support is very likely to affect user experience in a noticeable way.

Is is very difficult to find any HW without VP8 support. Android mandates it since 2.3, so since then the SoCs used in mobile devices support it.

Only Apple does not support it for political reasons. Just like they never supported Vorbis.

On the plus side, the situation is much less dire than it was back when Mozilla first made the argument against H.264. WebM is a serious alternative, and if MPEG LA tries anything funny, people will be able to switch.

Unfortunately, WebM isn't _really_ a serious alternative. It's a fine codec and produces great quality video in a similar filesize to H.264, but it's not supported on the majority of mobile devices. H.264 is also what many consumer video cameras produce, like iPhones and most DSLRs.

Pragmatically, I'm happy that all the major browsers (with the exception of Opera) will now support one common codec, even if it's patent-encumbered.

Depends on the threshold for calling something "serious". It's far more serious than any alternative before WebM, that's for sure. No competing format will ever be an instant drop in replacement for an entrenched format, but there is a continuum of scenarios below that.

From certain political points of view, absolutely. From a image quality point of view, not hardly.

The wrong choice is Flash.

H.264 vs. WebM is a far more nuanced tradeoff – widespread support and higher-quality implementations versus [hopefully] freedom from patents. Either way, the web wins if there's more <video> and less <embed> depending on an opaque binary blob from a single vendor with a track record of platform neglect.

Sure, but it could've been so much better, with almost no extra effort.

In the end, we have the worse of two options for absolutely no gain.

In the end patents will expire. 15 years left.

And that's one of the biggest problems with software patents. In 15 years, we'll likely be using a different codec (or, at least I would hope we'd make some progress in that amount of time).

Probably sooner, H.265 is next on the list http://en.wikipedia.org/wiki/High_Efficiency_Video_Coding http://www.h265.net

You say "no extra effort", as if widespread platform support, hardware acceleration, more than one high-quality implementation, significant support by authoring software, and direct-to-format hardware capture would happen for free. Avoiding royalties would be nice - although it's almost certain we'd still have lawsuits absent patent reform - but as a business decision that's a LOT to give up.

I really think Mozilla did the web a disservice. They should have implemented platform codec support immediately and added other formats so content vendors could adopt HTML5 video independent of the codec wars

How do they plan to defend against vulnerabilities in the system codecs? Especially a GPU-accelerated codec has the potential to wreak all sorts of nefariousness.

crosses fingers Please use libavcodec! Please use libavcodec! Please use libavcodec! reads article and is disappointed

By deferring to system codecs, they manage to defer some licensing questions, too.

It's a nice compromise, and it still encourages using royalty free codecs (since those are the only ones with a chance of guaranteed deployment, even if Apple is still acting funny these days)

they will use gstreamer (which won't rule ouy libavcodec).

gstreamer itself will use libavcodec

that's what i was implying (with my typo). but gstreamer might use something else as well.

Contrary to the article, h264 is not currently working on Android Firefox (2012-10-11 release), or at least it doesn't for me.

Reply on news.ycombinator.com