Minutes: MathML intent meeting, 21 October, 2021

Attendees:

   - David Carlisle
   - Sam Dooley
   - David Farmer
   - Deyan Ginev
   - Paul Libbrecht
   - Louis Maher
   - Bruce Miller
   - Murray Sargent
   - Moritz Schubotz
   - Neil Soiffer
   - Steve Noble
   - Cary Supalo
   - Bert Bos

Regrets:

   - Daniel O'Mahony
   - Moritz Schubotz

Announcements/updatesa

NS: Sad new -- Laurence Zaysser can no longer work with the group due to
her work commitments.

NS: If you can get MathML into pdf as fully tagged, it will be read by NVDA.

NS: V2 of PDF spec say you can have an associated file with MathML in it
tagging the <math> element that can make pdf readable. Framemaker is doing
this already, Adobe hasn't modified their accessibility interface yet to
allow AT to get at that info without a lot of extra effort. So sadly, at
the moment, no better off.

DC: Notes on the LaTeX project Tagged PDF project (with some Adobe support)
https://www.latex-project.org/publications/indexbytopic/pdf/. They are
leaning towards using associated files.
TPAC Breakout session post-mortem/feedback

NS: We had 30 people in the TPAC MathML meeting.

MS: Said NS did a great job.

PL: There is an extension point here. Is it documented? PL: If we use style
sheets then people can contribute their own ways to say things.

PL: If we have a way to input intent with CSS then there can be extensions
in the form of CSS style sheets.

PL: People could write style sheets that imply intent.

DC: Is was good to have that many people listen.

NS: Saw only one screen reader representative. She was an Orca
representative.

NS: Thought he was capable of putting his talking CAT software into Orca.

MS: Said Joanie liked the intent attribute.

BM: Separate the information model from the transport model.

SD: There is a separation from the information content and whatever is
delivered to the AT user. Jason White said this
https://github.com/w3c/mathml-docs/issues/36

DG: People liked simplicity. Everyone is stretched for resources. Keep this
simple and easy to implement.

NS: Wanted to have a discussion on the accessibility tree at the meeting,
but it didn't come up in the comments after his presentation.

Jason's coments From Sam Dooley to Everyone:
https://github.com/w3c/mathml-docs/issues/36

NS: People were OK with intent.

NS: This allows two approaches: include intent and either the screen reader
does something with it or use CSS. YOu can write CSS style sheets which
imply intent.

BM: The CSS would be providing intent data.

NS: CSS can use the structure and use selectors to generate intent, or
given intent, I can generate speech. We would use the selector mechanism of
CSS.

SD: Have user definable ways to set defaults.
SD: If we have information coming in from CSS and intent, we need a way to
decide which dominates.

DC: Finds CSS lite confusing. What is CSS lite?

DC: User Stylesheets may not be supported by some browsers.

PL: User stylesheets are useful for special needs, but they be inconflicts
withother sources of information.

PL: User style sheets would be nice to have.

DG: CSS is closer to the AT reader where Intent would be close to the
author.

BM: Is there any possibility to use an existing mechanism to do what math
needs. There was no response about this especially by the ARIA people. NS:
In order to keep things simple, he does not think ARIA will have any
changes in the speech area. NS: Jony says pushing things into ARIA label is
not good.

BM: If there is any positive response from screen reader developers to make
MathML less scary There was a hope that the math access could be provided
by AT developers and not browser developers.

MS: Generates speech for Narrator. Would like to have Narrator deal with
Math better. The UIA people do not want to deal with MathML.
Next step

NS: People said intent was OK. We should generate intent in MathML as
experiments. NS: will test the math with mathcat. Also would like to work
with ARIA label approaches via CSS.

NS: screen readers do not look at MathML or ARIA label.

DC: before we start writing intent, we should write a small spec on what
intent is. What is its syntax?

BM, DG BM and NS have worked with the syntax.

BM: Action will start looking at syntax.

BM: Where do unknown names go? ?Will they go into dictionaries?

NS: He is a fan of trying to implement something to see if your ideas pan
out. You must have some initial specification.

BM: ACTION: will write a one or two page to describe what intent is.

SD: take presentation + intent and generate content mathML.

BM: will need help from DG on this. May have something solid in a couple of
weeks.

SD: will send some information to BM to help with this.

DC: daylight savings time will change in a couple of weeks.


<http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail>
Virus-free.
www.avg.com
<http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail>
<#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>