[whatwg] <video> element proposal

[James Graham]

Håkon Wium Lie wrote:

> The main arguments against specifying formats is that the market
> will sort it out, and the standards would easily be outdated if
> they required specific formats. 

Arguably this is irrelevant. If a video element gets any significant 
adoption it will be impossible for any future browser to drop support 
for whatever format is widely deployed, even with the advent of superior 
technology. Indeed, the multi-client nature of the web means that in 
general it is quite conservative in what formats are adopted; for 
example JPEG2000 and MNG have failed to make any impact whatsoever 
despite the latter having had a major implementation -- they simply 
don't offer enough benefits over the existing technologies.

Therefore, once a <video> element is widespread enough to be important 
there are likely to be a small group of "good enough" formats which will 
have to be supported in perpetuity, even if better ones come along. 
Accepting this hypothesis, we should not worry about the possibility of 
the spec being outdated but the possibility of the spec being unadopted. 
What do we need to be adopted? A minimal criterion is to be as good as 
the market leader (currently Flash video, IMHO). A more realistic 
criterion would be "a significant advantage over the market leader" but 
let's focus on the easier case first. To be as good as flash we need:

Widespread deployment (despite not working in many situations Flash has 
close enough to 100% of the desktop market for content producers to 
ignore the rest). In practice this means at least two things. Firstly, 
browser manufacturers without significant in-house video expertise 
should be able to implement the spec using an external library. Secondly 
it means that we should be able to implement it in existing versions of 
IE somehow.

Interoperability between all implementations. If different 
implementations support different formats it's easier for everyone just 
to use Flash which works the same everywhere.

The same level of functionality for authors as Flash offers. This 
includes e.g. scripting interfaces for custom players. I don't know the 
details of this so I'll stop here.

IMHO, specifying a baseline format in the spec will make these meeting 
these first two requirements significantly easier. However the second 
requirement means the format has to be royalty free and preferably have 
a open source implementation available. This seems to imply that the 
spec should mandate support for at least Ogg Theora.

-- 
"The universe doesn't care what you believe. The wonderful thing about 
science is that it doesn't ask for your faith, it just asks for your 
eyes" --- http://xkcd.com/c154.html