Re: Text associations for HTML5 video/audio

Hi Silvia and Sean,

Here is the bottom line on the proposal that I made:
a) we should have a well-defined selection mechanism for alternative 
content that is scalable.
b) the processing should not include special-purpose processing for 
historical reasons, if possible.
c) it should be possible to reference embedded and external text tracks.

I think that it would be productive to talk thru these issue in a 
teleconference, rather that converse via text, since this introduces 
assumptions that may not be valid. I'm happy to this in a large or small 
group.

One of the things that has been bugging me most in thinking about this 
is that the addition of text associations into the <audio>/<video> 
object (whichever way we go) is actually a kludge: it is a way of 
avoiding temporal composition of text along with other media. Cramming 
this into the video element is actually not the right way to go: it is a 
stop-gap measure that has no real chances for extensibility. I would 
much rather look at a real solution to the composition problem, since 
this would much better need the needs of accessibility (and other) use 
cases.

-d.

Received on Thursday, 22 April 2010 14:34:17 UTC