Re: Timed tracks

Hi, Ian:

I'm quite concerned that WebSRT has been inserted into the HTML5 editor's draft (http://dev.w3.org/html5/spec/video.html#websrt ) when it hasn't been discussed in the accessibility sub-group.  There isn't any evidence, nor has there been any consensus as far as I know, that this format is appropriate or fulfills any requirements for captions or subtitles- requirements which are now being formulated by the group but are not yet final. My immediate opinion is that the introduction of WebSRT in this manner will only cause confusion:  readers may take its inclusion as an indicator that this format has W3C endorsement and is on an approval track, and therefore can be used in test implementations.  In a field that is already overrun with timed-text formats, this will only serve to confuse the situation further.  In fact, it could be very difficult to reverse later.  In the interest of not displacing or damaging the work being done by the accessibility sub-group, I think it would be best to remove all references to WebSRT from HTML5 documentation at this time.

Geoff Freed
WGBH/NCAM




On 5/5/10 5:10 PM, "Ian Hickson" <ian@hixie.ch> wrote:

On Wed, 5 May 2010, Philippe Le Hegaret wrote:
>
> while I appreciate your enthusiasm and time that you're spending on this
> issue, I'm surprised to see that you're incorporating a new captioning
> format in the HTML5 specification without discussing about it.

We've been discussing this for close to a year now.


> imho, the HTML5 specification shouldn't attempt to define a new
> captioning for video:
>  http://dev.w3.org/html5/spec/video.html#websrt

It's not a new format; it's the format used by SubRip, with minor
extensions to handle features that were found to be needed during
research (such as vertical text, ruby, and karaoke).

--
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'