[math-on-web] a11y TF meeting minutes, 2017-03-13

Hi everyone,

below are the minutes from the a11y task force meeting this week.

See you later for the regular call.
Peter.

# 2017-03-13 MathOnWeb a11y TF

* Peter: extended the demos, some more tests with MathML AT
  * Dani: summary would be great
    * Peter: blog post as soon as I get to it
* Charles: at CSUN spoke with Doug Schepers who wants to provide more
deep-data inside SVG so that other frontends can access that data
  * if it's just about accessibility, then ARIA makes sense
  * but if it's for other tech, then other it might need a different
technology
  * Dani: if it's not just math but phsyics or music
* Charles: right. STEM obvious target
  * but where does that fit in W3C
  * is there even anythig beyond ARIA
  * Peter: HTML maybe, but currently seems focused on interactive
(summry/details etc)
* Dani: we're mixing ideas a bit
  * native features vs new tools
  * I'm not expecting native features for math (a11y)
  * but tools that could add aria roles / labels etc
  * could be dependent on actual content, with JS
    * so authors don't worry about it
  * Peter: agreed. contextual information will be critical.
* Charles: dynamic makes sense
  * also language dependence
  * speech rules per teachers

* Peter: let's discuss role=math
  * Charles: we had issues with double voicing
    * Peter: implementations are even messier than the "dumb" role
  * Charles: but if we improve it, then physics, music, etc too?
  * Peter: right. I'm more for identifying neutral features.
  * Dani: agree. But what is role=math for then?
  * Peter: could still be useful as a fallback position
    * if ARIA objects for general feature, we can at least do it in math
  * Charles: at CSUN there was a project where an interpreter is reading
chunks
    * Peter: MathJax's a11y extensions do this to some degree
   * Charles: PHET group providing separate DOM tree for a11y to make their
interactive animations work
    * at CSUN hackathon, did some work on chunking this
* Daniel: what's next here?
  * Peter: gather some use cases for role=math, e.g., what comes from your
experiments
  * Daniel: don't really see the need; it's not a button or anything
special
  * Peter: but providing info like list, table, link
    * Charles: heard similarly from ETS
* Peter: perhaps best to focus on what helps make accessible solutions
possible
  * ACTION: Call for comments / next CG meeting