W3C home > Mailing lists > Public > w3c-wai-ua@w3.org > April to June 1999

Minutes from 19 May 1999 UA teleconference

From: Ian Jacobs <ij@w3.org>
Date: Wed, 19 May 1999 13:51:01 -0400
Message-ID: <3742FA05.703F9B58@w3.org>
To: w3c-wai-ua@w3.org
WAI User Agent Guidelines Teleconference
19 May 1999

Chair: Jon Gunderson (from Madison)
Scribe: Ian Jacobs

Marja Koivunen
Harvey Bingham
Chris Weaver (New Mexico State)
Madeleine Rothberg (WGBH)
Mark Novak



- (NEW) Action Chris: Write up Math techniques proposal and post to
                 list. MR to review this.

- (NEW) Action editors: Incorporate resolutions into next draft of

- IJ: Write DJW about requirements T&S/WAI.
  Wrote twice, no reply. Will follow up.
  (Continue All IJ actions)

- CMN: Write techniques for 7.2.2 and 7.2.6
  CMN deferred until publication of NOte by Rich and Mark.
  Probably continued.

- JG: Techniques for 7.2.2.
  Not done.

- JG: Techniques for 7.2.1.
  Not done.

- JG: Contact Rob about 7.2.4/5/6
  Rob hasn't been participating and we have consensus
  on checkpoints in this section. 
  JG: I need to contact Rob about this. We want
  TECHNIQUES that point to vendor-developed (internal)
  guidelines. This for platform-specific conventions.

- JG: Contact Rich S. and Peter Korn for techniques.
  JG: I contacted Rich but not Peter. 

- JA: Check guidelines for information about tooltip control.
  Not done.

- Editor: Group navigation by function.
  IJ: We may have been here already. Will look into how this 
  would fit into the document.

- MK: Propose navigation checkpoint to time-sensitive
  parts of a document.
  MK: Not done.
  JG: This is about navigation the SMIL tree or parts of 
  a multi-media presentation.
  MK: If a link is time-dependent, allow the user to navigate
  it independent of time.
  JG: Review this in terms of stop/start/rewind/fast-forward


- HB: Checkpoint for metadata search.
  HB: This became attribute-value search. But in addition,
  give user information about attribute/value pairs.
  IJ: Is this a good feature or a solution that requires users
     to know even more about the document?
  JB: HB should write a technique that shows how it's useful
      for accessibility.
  HB: Already done.

- JG: Add issue to issues list about "accesskey" recognition.
  JG: I think I did that. On the mailing list, the conclusion
  to proposals was not to have a checkpoint about support
  for access keys. Instead, discuss it in HTML accessibility 
  features section. Implementation details
  in the techniques document.

- JG: Organize Math discussion.
  Done. Had discussion last week. On agenda this week.

/* Brief discussion of presence of accessibility at WWW8 */

   Brief discussion of difficulty of getting UA Guidelines
   to Recommendation 

HB: This is where the rubber hits the road.

JG: People will buy into them when they start asking their tools
    for support.

Agenda 1) Discusion and resolution of support of math
Issue #12: Recommend full implmentation of MathML http://cmos-

JG: Three main conclusions:
  a) Provide access to math document tree (preferably MathML).
     Add this to current navigation techniques.
  b) Recommend MathML if user agent/plug-in handles math.
  c) Make math information available to assistive technologies
     (e.g., XML tree).

MR: The alternative to MathML is that authors create graphical
    images of math, which is problematic. Can't we encourage
    MathML more strongly? 

IJ: Checkpoint that says "If you do math, use MathML" and motivate
    this instead of requiring authors to use images.

CW: What do we mean by "rendering" math?

IJ: I think that we mean prefer MathML as a markup language.

CW: What if a UA encounters a math document on the Web. Should
    we recommend that UAs be able to read this, or is more
    of the domain of plug-ins?

JG: In Web Content, this was discussed and conclusion was that
    math and science are part of a larger framework.

IJ: The larger framework is "Use W3C Recommendations where available."
    Maybe need a mirror checkpoint to "Implement applicable W3C

JG: Say this in the Techniques document? 

IJ: I think this can be said in guideline themselves.
    E.g., may want to encourage support for deprecated features.

PROPOSAL: Review WCAG checkpoints. See if this group covers all
    of them or needs to mirror them. Checkpoint 11.1, if mirrored
    by UAGL, would cover MathML (which could be listed as an

RESOLVED: Bring WCAG checkpoints from guideline 11 into
    Guideline 7.1. In short, rework 7.1 based on that.
    Ensure that Math is part of that as an example. SMIL
    as well.

CW: Is a similar concern being taken to the AU WG?

JB: They will probably reference UAGL in this area.

Agenda 2) Issue #13: Provide keyboard access to MathML similar to forms

JG: What's between navigation of math document tree and navigation
    of linear navigation of equation.

IJ: What about lower bound/upper bound of an integral?

JG: Linear is appropriate for simple equations. Need this for simple
    situations and naive users. Discussion last week about some
    middle ground. Maybe just talk about these two approaches
    in techniques; this might be sufficient for math in this document.

CW: You need to get through a math tree unambiguously. Linear is
   depth-first. Tree is a "node-announcing breadth-first search".
   Users should be able to shrink or expand subtrees. Need clues
   about what the children are (math nodes tend to be very sparse).
   You may need to know about the depth of subtrees in order to
   guess where to navigate.

IJ: I see the same techniques useful for, e.g., navigating
    a document structure such as headers.

CW: Math is a somewhat special case due to sparseness. We should  
   be sure that the guidelines ensure that the tree structure
   is apparent. Yes, it fits under a checkpoint about navigating
   the document tree.

CONCLUSION: Just include math in current checkpoints.

Action CW: Write math technique proposal (with others if desired,
    e.g., Raman and Gardner).

Action MR: Review this proposal.

Agenda 3) Issue #14: MathML objects should be included as part of the
document tree

JG: Ensure that we mention XML in checkpoint about exposing tree.

IJ: Do we need to say anything specific about math if it's covered
    in sections on (1) Use W3C specs (2) Use the DOM?

HB: Need to mention entities/mapping to semantics.
   User agent issue to
    convert symbols into (notably spoken) rendering that is
    part of the math terminology.
IJ: To me, I hear "semantics" and I think "schemas." 

JB: Do we need to say "support schemas"?

IJ: Political issue right now about convergence of RDF and XML.

CW: If the author has made the spoken rendering available (e.g.,
    through a schema), then the user agent should make that
    information available to the user.

MK: What about the reading order? 

JG: Is this a navigation issue? Or do we need to say something
   about "when markup is available for content substitution, make
   the content available".

IJ: Covered by 5.2.1. Perhaps ensure that namespaces included in

CW: MathML provides separate content markup for use by computation
    engines. It's supposed to be included with the math and can
    be passed around, e.g., for evaluation of expressions.

JG: Put details in techniques doc. 

Agenda 4) Structural/Editorial changes to document (based on 
     Web Content Accessibility Guidelines experience)

RESOLVED: Adopt single-digit numbering of guidelines as per
    WCAG. Push "structure" to introduction.

CW: I like single-digit number systems. They should agree
    across guidelines for consistency. 

IJ: Will be difficult to track changes in WCAG. I think this
    WG should choose the most important and put them first.

RESOLVED: Move terms/definitions to a glossary at the end.
Received on Wednesday, 19 May 1999 13:51:32 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 October 2009 06:48:57 GMT