[whatwg] Video Tag Proposal
Sir Gallantmon (ニール・ゴンパ)
ngompa13 at gmail.com
Sun Mar 28 11:29:52 PDT 2010
On Sun, Mar 28, 2010 at 10:49 AM, Ashley Sheridan
<ash at ashleysheridan.co.uk>wrote:
> On Sun, 2010-03-28 at 10:29 -0700, Kelly Clowers wrote:
> 2010/3/28 Sir Gallantmon (ニール・ゴンパ) <ngompa13 at gmail.com>:
> > When the img tag was made, all browsers initially supported BMPs, didn't
> > they? Nobody complained about implementing support for an image format. The
> > GIF format made things hairy later, but with JPEG and PNG, the issues
> > eventually resolved themselves. But the img tag was made at a time when
> > there was no format soup for images... Or at least, not one nearly as
> > serious a problem as the video tag.
> Also, image formats are less complicated to implement than video, both in
> the codec itself and the fact that images don't have to worry about subtitles,
> containers, seeking, sound, etc.
> > Without a baseline codec, there is no guaranteed usefulness to the audio or
> > video tags. As for audio, I suggest supporting at least WAV (or FLAC) and
> > Vorbis at least.
> That was the recommendation before all codec references where removed.
> Currently among HTML5 browsers, I believe only Safari does not support Vorbis
> (they all support pcm wav). Safari uses QuickTime, so if Apple would bundle
> XiphQT, Safari would be set. Who knows what MS is planning for IE9, but I don't
> think they would object too much to having Vorbis as an option,
> especially if they
> are using DirectShow. A great many high-profile games have used Vorbis,
> including MS-published Halo and Fable.
> Really, the audio situation seems fairly manageable. Vorbis even has an
> advantage in size/quality over most other codecs, especially the so-common
> > For video, our best shot is either Dirac or Theora. Unless
> > somebody else has any other decent reasonably available open source,
> > royalty-free codec that can be used for the video and audio tags?
> Well, if Google frees VP8...
> Kelly Clowers
> I was under the impression that Apple were one of the main opposers to
> using free codecs in-place of their proprietary QuickTime.
> Also, when was the last time you ever knew Microsoft to go with standarised
> formats when they can just as easily push one of their own?
> Even the image formats in the early days were an area of debate. Does
> anyone remember the time when there was still the threat of the licensing
> issue surrounding the Gif format? It was this very issue that cause the PNG
> format to be created.
I remember the Unisys LZW patent issue. The problem is that the situation
surrounding GIF at the time and the situation we have now is totally
different. Technically, LZW compression didn't have to be used with GIFs,
though that practically didn't happen. And now we have a situation that
companies like Apple and Microsoft themselves created, with the crazy
redundant and overly broad patents. None of the companies on the MPEG-LA
want to fix the situation because over half of them only make money through
Additionally, most patents covering the video formats are unfortunately core
to the implementation of the codecs. As an earlier mail said, video codecs
are hard to design, much less implement. Video codec companies that spend a
lot of R&D time do deserve to make money on their hard work. However, when
every single codec is in trouble because of overly broad patents, we have
I've said it earlier that I'd like to have some baseline codecs specified.
Can we at least add Vorbis and WAV (and/or FLAC) to the baseline codecs for
the audio tag? Nobody has really complained about it, so why not put that
back? At least then, one issue with codecs is resolved.
Finally, spending time on if's is not going to do anything. For the purposes
of these discussions should assume that Google will not open up VP8 or any
other codec owned by On2. We need to work with what we have NOW, not what we
MAY have LATER. If Google announces tomorrow that they open up VP8 in the
same way that Theora was, then great! But if we focus on if's like that,
we'll never get anywhere.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the whatwg