Media Subteam Teleconference Minutes for 29 June

Minutes from the 29 June teleconference of the HTML-A11Y Task Force
Media Subteam are provided below as text and are available as hypertext
at:

http://www.w3.org/2011/06/29-html-a11y-minutes.html


   W3C

                                                           - DRAFT -

                                          HTML Accessibility Task Force Teleconference

29 Jun 2011

   See also: IRC log

Attendees

   Present
   Regrets
   Chair
          John_Foliot

   Scribe
          Janina_Sajka

Contents

     * Topics
         1. Hierarchical Navigation: Progress Checkin
         2. Paused Media: Exploring the Optimal User Experience
     * Summary of Action Items
     __________________________________________________________________________________________________________________

   <trackbot> Date: 29 June 2011

   <scribe> Meeting: HTML-A11Y telecon

   <scribe> Scribe: Janina_Sajka

   <scribe> agenda: this

   <scribe> Meeting: HTML-A11Y telecon

   <scribe> Chair: John_Foliot

   <scribe> Scribe: Janina_Sajka

   <scribe> agenda: this

   scribenic: janina

   <Sean> +44.154.558.aaaa is Sean

Hierarchical Navigation: Progress Checkin

   <silvia> http://www.w3.org/WAI/PF/HTML/wiki/Media_Navigation

   sp: Suggestions for attaining DAISY type navigation in HTML 5 media ...
   ... driven by external text tracks
   ... Have not thought through all the issues on the AT side, but believe what we have is sufficient for support

   js: So, if we have media resource and two alternatives, audio based video description pluse a sign translation video,
   we need to create a text track to get navigation?

   sp: yes

   <JF> <ul>

   <JF> <li><? target='timestamp' data='00:00:47.600'>Impact of Captions on the Web</li>

   <JF> <li><? target='timestamp' data='00:01:50.100'>Requirements of a Video text format</li>

   <JF> <li><? Target='timestamp' data='00:02:30.000'>The Web VTT Format

   <JF> <ul>

   <JF> <li><? target='timestamp'

   <JF> data='00:03:33.000'>Simple WebVTT file</li>

   <JF> <li><? target='timestamp'

   js: do we need a new label for that otherwise empty text?

   <JF> data='00:04:57.766'>Styled WebVTT file</li>

   <JF> <li><? target='timestamp'

   <JF> data='00:06:16.666'>Internationalized WebVTT file</li>

   <JF> </ul>

   sp: No, it's called "Chapter."

   <JF> </li>

   <JF> </ul>"

   sp: JF's question ml looks correct ...

   jf: like fly-out menus
   ... what about two text tracks with different levels ...

   sp: believe same ml will work for ttml as i've shown with webvtt
   ... We've both vtt and ttml in the wiki page--both can map
   ... Sean, were you able to look at the wiki on navigation?

   sh: looked at ttml, looks workable

   jf: just concerned that our ml is acceptable both for vtt and ttml authoring

   sh: It comes down to conventions, so yes. This is a perfectly valid way of doing it.

   fo: are we defining a new collection of objects? are we overloading queues? How do we think of handling this from js?

   sp: Only thing I've introduced a nest element, start and end, for grouping
   ... It only displays as the time is reached on the timeline

   fo: So, this is a new concept. So, today, you can load queue from external file, or querry. How does chapter concept
   fit in?

   sp: I'm not introducing anything new.

   fo: How do I skip from one chapter to the next?

   sp: I haven't addressed that side in this wiki

   fo: I think you misunderstand my question
   ... You can load from file queus, manipulate with js ---

   sp: Is the same
   ... All I do is used a text track that's associated with video as "Chapter" kind, and provide a queue ist.
   ... I'm showing it's contents

   <silvia> http://www.whatwg.org/specs/web-apps/current-work/multipage/the-iframe-element.html#texttrackcue

   sp: I picked this object and filled in all the attributes

   sh: so the chapter points are distributed throughout ...

   sp: top levels are separate queues

   sh: sublevels are all separate
   ... doesn't seem elegant

   sp: it's what DAISY does

   <silvia> http://www.html5videoguide.net/demos/google_io/3_navigation/

   sp: DAISY has one structure, this is two

   jf: sounds like chapter notion is abandoned, these are just specific time queues

   sp: yes, because original "chapter" concept was DVD based, and this is independent of that. This is a different
   semantec notion.

   sh: Not sure chapter was defined to be equivalent to DVD markers?

   <silvia> "Chapter titles, intended to be used for navigating the
   http://www.whatwg.org/specs/web-apps/current-work/multipage/the-iframe-element.html#media-resource. Displayed as an
   interactive list in the user agent's interface."

   <silvia> http://www.whatwg.org/specs/web-apps/current-work/multipage/the-iframe-element.html#the-track-element <-
   scroll to the list

   jf: So "chapter" was a poor term to choose.
   ... Perhaps we should suggest a different term.

   sp: will always be someone who misunderstands, no matter what we call it

   sh: there's no implication that chapters can't be nested

   jf: just as with <li>, there was an implicit </li> when the next<li> was presented?

   sh: but there could be another chapter inside, at a internal nesting level

   sp: I think the nesting is more obvious the way I have it
   ... It's true that we could organize this differently.

   <silvia> "<? target='timestamp' data='00:00:47.600'>"

   js: Bottom line here, we need nothing new from HTML spec to do hierarchical navigation? Am I wrong?
   ... We were working on a July 1 deadline

   <Sean> <li><? target='timestamp' data='00:00:47.600'>Impact of Captions on the Web</li>

   sh: My question is how this is to be exposed into html

   <silvia> http://www.whatwg.org/specs/web-apps/current-work/webvtt.html#webvtt-timestamp-object

   sp: through xml processing instructions

   fo: would be very very good if we can keep our existing queue structure the same. It would be unfortunate if we need to
   change that.

   sp: There's no change

   fo: Think we have to do something to expose chapter info
   ... We need some discussion on how to do chapter navigation programaticallysp: I can show the same working with DAISY
   or TTML, inside queues. It can end up in the same ml

   sp: Only objection I hear that is a problem is that the drilling in navigation is inside queues
   ... reason is that representing it on a timeline would be hard, and I think we don't need that

   sh: Didn't et to my point, there's a ? inside the nesting, not sure what that is in the html
   ... Not sure what that is

   sp: It's an xml processing instruction that already exists in html
   ... Now looking for a link pointer

   sh: I understand it in vtt, but not in html

   <silvia> http://www.whatwg.org/specs/web-apps/current-work/webvtt.html#webvtt-cue-text-dom-construction-rules

   sp: I'm mapping the nest item into an ordered list

   jf: Want to focus on any cp we need to deliver by friday?

   janina: also wanted to say that, but also ...
   ... showing hierarchical nav points may not work on a timeline, but it can work visually very well, perhaps in a system
   map kind of display

   jf: Yes, and nested lists is a familiar concept for web authors.
   ... Would this xml be the cp?

   sh: Would be a very interesting proposal as other xml has been frowned on in html

   sp: Only proposal here is related to webvtt, so from that point of view we don't need a cp
   ... I could show how DAISY could be mapped in, and also ttml.

   sh: When you have the get queue as html can you remove the xml?

   sp: it is html

   sh: no
   ... xml processing is suggested to be ignored by dev html

   <silvia> http://dev.w3.org/html5/spec/Overview.html

   <silvia> ProcessingInstruction

   sh: I see no need because of this, nothing is different

   sp: it's all part of the dom, text, event, etc
   ... Perhaps an example into the spec of how the hierarchical ml would work, an informative section

   jf: If we get this xml processing piece to a point where there's agreement, are you otherwise comfortable with this?

   jb: Best that we keep this, so putting a proposal means there will be a call for counter proposals, and from that we
   can work toward the best approach.
   ... How much to get this done?

   sp: It's due Friday
   ... I suggest I forward this myself, as a cp, since we don't have agreement in the group. That would take a couple
   hours.

   jf: People were saying it would be good to have examples in addition to vtt

   sp: Yes, will do.

Paused Media: Exploring the Optimal User Experience

   jf: I think aria-described won't work, fundamentally, unfortunately.
   ... It's because it's off page so can't focus and links and ml would be lost, because they're outside of the viewport
   ... it's all about tab focus

   <silvia> aria-describedby is not off-page

   jf: then we don't meet the design

   leonie: every designer will hate it

   sp: for that we've ... [[missed]
   ... it's inside the video element, so it's fallback, not visually presented, but in the dom

   jf: can't have it in the dom and not available to sighted users

   <JF> <video>

   <JF> <p id="posterdescription">This <a href=""></a>link</a> is a tab-focus item</p>

   <JF> </video>

   <judy> [Judy sends regrets for next week; will be away]

   jf: Silvia will submit a cp on nav, we'll keep discussing it and other approaches. And, we'll keep working on poster.

Summary of Action Items

   [End of minutes]
     __________________________________________________________________________________________________________________

-- 

Janina Sajka,	Phone:	+1.443.300.2200
		sip:janina@asterisk.rednote.net

Chair, Open Accessibility	janina@a11y.org	
Linux Foundation		http://a11y.org

Chair, Protocols & Formats
Web Accessibility Initiative	http://www.w3.org/wai/pf
World Wide Web Consortium (W3C)

Received on Wednesday, 29 June 2011 23:16:27 UTC