Raw minutes from 22 June UA Guidelines teleconf

22 June 2000 UA Guidelines Teleconference

Present:
 Ian Jacobs (Chair/Scribe)
 Harvey Bingham
 Kitch Barnicle
 Eric Hansen
 Dick Brown
 Gregory Rosmaita
 Tim Lacy 
 Charles McCathieNevile
 Mickey Quenzer
 David Poehlman

Regrets:

 Jon Gunderson
 Mark Novak
 Jim Allan

Next meeting: 29 June.
  Regrets: Eric (probably for next month), Jon, Kitch
  
Agenda [1]
[1]
http://lists.w3.org/Archives/Public/w3c-wai-ua/2000AprJun/0505.html

0) Review of action items:

  TL: I talked to Rob Rylea about independent volume
      control of audio sources. He said, that depends on
      what software is configured to be the default player.
      For IE, windows media player typically used.

  IJ: Could windows media player be used for each source of
      sound?

  TL: You can choose the player independently of the default.
      I don't know what happens when you have several media
      players going at once. 

  (New) Action GR: Pursue speech synth ranges for other 
        properties than speech rate.

2) PR#287: Proposed clarification to checkpoints 3.3, 3.5, 3.6
   http://cmos-eng.rehab.uiuc.edu/ua-issues/issues-linear.html#287

   In particular, this proposal:
   http://lists.w3.org/Archives/Public/w3c-wai-ua/2000AprJun/0496.html 

   Refer also to HB's proposed modification:
   http://lists.w3.org/Archives/Public/w3c-wai-ua/2000AprJun/0504.html

   GR: I think we want to separate content that is blinking/marquee.
       Red flag: is this on a piece-by-piece basis or globally.

   HB: Is it consistent with marquee that it always moves in a
       reading direction?

   GR: Not sure for non-standard markup.

   TL: The MARQUEE element supports directionality.

   IJ: I think the issue was movement, do we need the fine
       control?

   EH: But if you freeze everything, it would make it hard
       to interact with the page. Presumably, a global control
       would freeze all video in the document.

   KB: Is restart the same as removing pause?

   EH: "resume"?

   Consensus on "resume".

   HB: In SMIL, you can jump to anchors in video.

   IJ: One proposal is to treat video and audio
       individually,  but animations and blinking 
       globally. But of course, you could treat
       synchronized components as a unit.

   EH: Isn't the focus generally on one particular
       video, rather than multiple video?

   KB: We've talked about a video track of sign
       language.

   EH: But presumably they would be linked so that
       they could be controlled in sync.

   KB: Yes, in that situation. But you could imagine
       three sports matches on the screen at the
       same time.
 
   EH: It's common to find several animated graphics
       operating independently on the screen.

   IJ: Maybe we can make a distinction based on
       "importance". Video seems to be more
       important in general. 

   EH: One way to distinguish animations and
       video (without considering importance)
      is to say "start, stop, pause, and resume
      multimedia presentations and audio 
      presentations". I'm not sure whether 
      blinking content should be considered separately.      
      That way, in the future if the dividing line
      between animations and video blurs, we're 
      covered. This may be a case where global
      and independent control are useful.

   IJ: Like global and independent volume control.

   TL: (Back on the multimedia question):
       I have three players operating indendently on 
       separate feeds, and I can control them
       separately. If you had a Web page with 2-3
       embedded AVIs, and they were set to play
       on load and to use a plug-in, it should work.
      

   IJ: That's great, except that some sources may
       be not handled by the UA (but other software).

   TL: Also, they can run asychronously. Suppose you
       had 2 AVIs using HTML+Time to display dynamically
       changing captions. What would happen: as each
       caption changes independent of each other, it would
       get a focus event, meaning that the screen reader
       would jump between them.

   GR: When HPR first came out, this was a problem - two 
       instance of screen readers. 

   GR: With pure audio, UI is spoken by screen reader and
       AVIs are left alone.

   DP: How do you read the text in a SMIL presentation.

   MQ: It depends on the browser. WebSpeak has problems with
       this (you can't have an audio stream going and browse
       at the same time). With Jaws and IE5, the access program
       is monitoring the UI that got spawned by the presentation.
       The browser would still be active, but if it didn't have
       focus, you wouldn't know what was going on.
 
   GR: Also UI issues - when I was listening to LPPlayer with
       Jaws. 

   IJ: How usable is the interface where there are three
       media players going at once? 

   TL: About the same as a picture in a picture on a TV.
       There could be some utility to this.

   TL: What about, if you have 3 streams, you need to be
       able to turn two off?

   IJ: That sounds like it's not good enough: you may need
       to hear all of them at once, at different levels.
 
   IJ: I think that default level should be global, and
       that control is available individually (on a 
       page-by-page basis).

   IJ: Any objections to the proposal:
      
http://lists.w3.org/Archives/Public/w3c-wai-ua/2000AprJun/0434.html

       1) Allow the user to configure the global audio volume
          (including silent). 
          Note: Do this at the OS level or the assistive technology
                level.

       2) Allow the user to control the volume of audio sources
          independently.

       Delete 3.2, 3.4

   CMN: How far to you have to be able to turn up the volume
        before it becomes a hardware problem?

   IJ: I would say that applicability kicks in at that point.
 
   Consensus: Adopt this proposal.
   Action IJ: Incorporate in the next draft.

   EH: For the other proposal:

     a) Handle multimedia presentation, audio presentation (audio
        only) and stand-alone visual track.

     b) For all of those, advance, rewind, start, stop, pause, and
        resume.

     c) For blinking content, one idea was to be able to override
        the blinking (perhaps toggling between on and off state).
        IJ: Is this pause and resume?
        EH: If it's pause, is the content off or on?
        IJ: For me, the minimal requirement is stop and have
            apparent content. The ability to restart is not
            as apparent.
        GR: Are we talking about just text?
        IJ: No, I think content in general.
        EH: Is blinking content either a standalone visual track?
            If so, it falls under category "b". If we're talking
            about a specific case for HTML, we could limit ourselves
            to the blinking text case.

   /* Mickey drops off */

        EH: Maybe, if we have fine-level control for the
            items in group "a", maybe we can concentrate on
            blinking text only. 

        IJ: Your proposal drops the blinking image case.
            
        GR: How do UAs know when something is animated?

        CMN: The image contains some control code.  Rollovers
             are done with javascript. Browsers allow you
             to stop gif animations. There's no technical
             barrier to having that as a default behavior.

         IJ: The requirement to step-through animation is
            P1 for access to all content.

      Proposal:
        1)  Advance, rewind, start, stop, pause, and resume
            multimedia presentation, audio presentation (audio
            only) and stand-alone visual track.

        2) Freeze blinking text and render statically.

        GR: What about scrolling text?

        IJ: Is that an animation?

        EH: There are cases where an animation looks exactly
            like text. Do blinking or scrolling text constitute
            animations?

        HB: Captioning would be another example.

        IJ: Yes, how do captions fit into this?

        EH: Maybe we don't need a special checkpoint for text.
            Maybe mention this in a note: include this
            as a "stand-alone" visual track.

        GR: I think there's a distinction between visual
            track and text, because you could want all text
            at once. 
   
        EH: Then we would exclude moving text from #1
            and indicate that it's covered in #2.

        Refer to Eric's background on multimedia:
       
http://lists.w3.org/Archives/Public/w3c-wai-ua/2000AprJun/0503.html

        Action IJ: Propose multimedia definitions to the list
        to define borders of these proposed checkpoints. 

        EH: One question: Should a standalone visual track be
        considered a special case of a multimedia presentation?

        EH: One idea: allow the user to configure the UA for
            consent before playing an auditory track.

        IJ: Why is this different from blinking or moving content?
            I could see this being done globally, but I wouldn't want
            to have to confirm the playing of each source.
         
       CMN: (On the question of what the viewport is for sound):
            Does real stereo mean different viewports? I think the
            idea of one speaker is insufficient. 

        EH: We have tracks, channels, viewports, and speakers.

       CMN: We address issues like positioning of multiple viewports
            and synchronization of viewports. When you have multiple
            channels at once, you need to be able to configure how
            you receive those channels.

        EH: Do we have checkpoints that cover GR's requirements
            already? One requirement would be "Be able to configure
            the UA to prompt before opening an additional viewport."
            Wouldn't this cover the case of launching an application
            that plays audio? Or maybe we go back to what I suggested
            earlier: promt the user before playing a new audio
presentation
            or standalone audio. You could have a similar checkpoint
            for video as well. Should this be a universal requirement:
            before you start a presentation, you should be able
            to configure the UA to prompt before opening.

        IJ: I don't want to be prompted for each loaded image.

        GR: I continue to return to the analogy of inadvertent
            submission of forms.

        IJ: I agree that the "on/off" requirements haven't been
            clear about whether this means globally or on
            a source-by-source basis.

        EH: What about 4.15: control of focus changes: Would this
            cover loading of audio?

        Action GR: Re-examine the orientation checkpoints and
        see whether they can be clarified to account for control
        of rendering of audio (and possibly other content) on load.

        EH: We're focusing on audio on-load. What about prompting
        for individual graphics or animations? 

        DP: I'm not sure what the accessibility issue would be 
            for other content types. I can imagine cases where
            we might want to load background images and sometimes
            we might not.

        GR: Turning images on or off is a conscious decision not
            to load images. I don't think you should be prompted
            on an image-by-image basis. I'm more concerned about
            those things that are not static.

-- 
Ian Jacobs (jacobs@w3.org)   http://www.w3.org/People/Jacobs
Tel:                         +1 831 457-2842
Cell:                        +1 917 450-8783

Received on Thursday, 22 June 2000 15:38:48 UTC