Re: [whatwg] Video codec requirements changed

Hello Chris,

I wanted to share my thoughts about this codec discussion with you, as I am
getting slightly worried about the turn the discussion takes. I joined this
mailing list several months ago, as one of the "invited experts" mainly to
follow the discussion about media support in HTML5.

It seems that several implementors have been advocacing very strongly
against a "MUST" support for any opensource and patent-free video codec, and
you just rejected the use of theora. If we fail to define one, or several
accepted formats to publish video on the web, I am afraid that every
implementor will go into a war to push its own. We will see all windows
live!  blogs starting to use the <video> tag with wmv and Microsoft
implement only wmv in their frontpage products, all mac users use only use
mov files... and several platforms won't be able to play these videos
because the implementor deliberately decided not to make a codec for them on
that specific platform. In other words: the web won't be fully
interoperable, many platforms not being able to access some of its content.

Having at least one codec that MUST be supported by the browser is the only
way to make it possible to publish video content available to all platforms.

But, even if the browser support the codec that it has to, I am also afraid
that many companies will still start to push their own formats, eventually
resulting in a "de-facto" standard which is not the one, or one of the
defined by the W3C. This has happened in the past with the use of the
patented GIF format for example, even if it is now free of patents - the web
has been using patented picture file-formats for years due to the absence of
specification, assuming that GIF and JPEG would be displayed on most
browsers. This was clearly a very bad way to go.

 In fact the standard could not be respected at all, and the page would
still be fully valid HTML5 -  leading to the balkanization I described
above. Let's not reproduce mistakes that have already been made in the past.

I therefore ask to add a section to the <video> section stating that the use
of non-w3c approved video codec in HTML would make the page invalid.

Concerning the previous debate about which standard to use, to be honest I
am not a video codec expert, neither am I really fully aware of the patents
risk, being european. I don't think I would be very wrong  though stating
that these points are very important:
- the format is able to be implemented on *EVERY* platform - and therefore
is open;
- the format is free  to use

This would make it possible for any web developer to integrate video in his
web page, whatever the platform he is working on, and make it possible for
anybody to read that page, thus enabling full interoperability - which this
is all about, isn't it?

I apologize for writing such a long mail to state something that probably
seems obvious to most of you, I just felt it was really important.

Thanks,

Yann Hamon






On Jan 12, 2008 12:02 AM, Chris Wilson <Chris.Wilson@microsoft.com> wrote:

>
> David Gerard wrote:
> >I have a suggestion:
> >
> >"Nokia, Apple: you want H.264, you free H.264. Make it irrevocably
> >perpetually royalty-free, it goes in. Do that with any other codec
> >that's technically better than Ogg Theora, it goes in. You can't do
> >that, we name Ogg Theora as a SHOULD. OK with you?"
>
> David, the problem is neither Apple nor Nokia (nor Microsoft, nor any
> other single entity of which I'm aware) can make H.264 (or MVC-1, or one
> of the variants of those) "free".  Perhaps the MPEG LA (
> http://www.mpegla.com/) could, but that's sort of like saying Apple could
> solve nearly all complaints against DRM in iTunes by making all the tracks
> free.
>
> On the other hand, anyone with a background in video and a clear
> understanding of the patent landscape appears to be saying Ogg Theora is
> likely to be subjected to a patent thicket similar to MPEG-2, they're just
> currently submarines waiting for a good target.  This is very unfortunate,
> because it sinks the idea that there's a single implementable-everywhere
> modern video codec, but that's a sad reality of software patents.
>
> In case I haven't been clear on this point - I think the <video> element
> is goodness.  However, I'm against even putting a SHOULD for Ogg Theora in
> the HTML5 spec.  It's not implementable.
>
> -Chris Wilson
>
>