Re: Systematic access to media/plugin metadata

On 12 April 2011 22:46, Eric Carlson <eric.carlson@apple.com> wrote:

>  My point was that as far as I know, no media file format has a standard way to mark a text track as a "transcript". Therefore the media format specific metadata has to be defined, eg. what 'udta' or 'meta' value signals a transcript in a QuickTime/MPEG-4 file, before the HTML WG can define a mapping to it.

But the HTML WG doesn't need to define mappings of specific fields to
specific formats, just the *interface* to use should the field be
available. Right now we have [1]:

[[
   A label
   This is a human-readable string intended to identify the track for the user.
]]

We could have:

[[
   A transcript
   This is a text representation of the spoken word in the media.
]]

- what either of those is mapped to in the media format-specific
metadata is up to 3rd party decoder developers, not the HTML WG.

Having said that, I don't personally believe it's desirable to specify
and/or restrict the naming of the potential name-value fields, better
to provide a general-purpose slot and "let the market decide".

Cheers,
Danny.

[1] http://dev.w3.org/html5/spec/Overview.html#text-track-label

-- 
http://danny.ayers.name

Received on Tuesday, 12 April 2011 22:43:17 UTC