RE: [minutes] Media sub-team teleconference: Oct. 27, 2010

IMO the simplest mechanism is to require that the HTML DocumentFragment obtainable from TimedTrackCue.getCueAsHTML(); be required to have CSSStyleDeclaration on the HTML elements to describe if/how it is to be rendered. It would then be up to the track format to define how to create these styled DocumentFragments (as I have proposed for TTML - see attached).

Then the positioning can be reduced to:

1.If the media element is a playback mechanism with no rendering area, abort these steps. There is nothing to render.
2.Let video be the media element or other playback mechanism
3. let textArea be a CSS containing block whose containing block is the rendering area for video, set the writing mode (CSS3) for textArea to lr-bt.
4.Let tracks be the subset of video's list of timed tracks that have as their rules for updating the timed track rendering these rules, and whose timed track mode is showing.
5.Let cues be an empty list of timed track cues.
6.For each track in tracks, append to cues all the cues computed as above for each each TTML cue event time.
7.For each timed track cue that is active, run the following substeps:
      1.Let nodes be the HTML fragment computed for the cue.
      2.Apply the terms of the CSS specifications to nodes to obtain a set CSS boxes relative to the CSS box created for the root div element in the HTML fragment, which is in turn positioned relative to textArea: [CSS].
     3.Add the CSS boxes in boxes to display
8. For each timed track cue that becomes inactive, remove the CSS box tree for its root div element from textArea:
   
This requires almost no additional specification in the HTML spec (pushes the responsibility into the TT format mapping), and allows the TT to address the full CSS space (so allowing captions and transcripts for example to be directly over, directly below, to the left of, or anywhere else in relation to the video's CSS box).

I'm fine with removing the CueSettings attribute I left it as a hook for formats that need it, but I agree that seems a bit kludgy. TTML wouldn't need it.

-----Original Message-----
From: Eric Carlson [mailto:eric.carlson@apple.com] 
Sent: 03 November 2010 04:22
To: Silvia Pfeiffer
Cc: Sean Hayes; HTML Accessibility Task Force
Subject: Re: [minutes] Media sub-team teleconference: Oct. 27, 2010


On Nov 2, 2010, at 9:01 PM, Silvia Pfeiffer wrote:

> I think we need to find a solution for how to add the positioning 
> rules on cues. Maybe we can go with existing CSS technology? Could CSS 
> settings be used here, just like on a <div> element? Do we even need a 
> CueSettings attribute?
> 
  I agree that we need to figure describe cue position before this is added to the spec. CSS seems like the logical way to do it, though I don't have any concrete suggestions.

eric



> 
> 
> On Wed, Nov 3, 2010 at 2:35 AM, Sean Hayes <Sean.Hayes@microsoft.com> wrote:
>> I'm aware I didn't define CueSettings, I'm not sure what the convention for that would be, so I'll get to that later..
>> 
>> TimedtrackCue at the end is a repeat and can be ignored.
>> 
>> -----Original Message-----
>> From: Eric Carlson [mailto:eric.carlson@apple.com]
>> Sent: 02 November 2010 15:21
>> To: Sean Hayes
>> Cc: HTML Accessibility Task Force
>> Subject: Re: [minutes] Media sub-team teleconference: Oct. 27, 2010
>> 
>> Hi Sean -
>> 
>> On Nov 2, 2010, at 7:51 AM, Sean Hayes wrote:
>> 
>>> In fulfillment of my action item, attached are the proposed edits I believe are required to make the current HTML5 technology neutral wrt WebSRT.
>>> 
>>  The "TimedTrackCue" section at the end of the document is empty, and you don't define "CueSettings". Did you send the correct version of the document?
>> 
>> eric
>> 
>> 
>> 
>> 
>>