Minutes: MathML general meeting, 1 July

 Apologies for the late minutes -- I thought I had sent them out the day
after the meeting.
Attendees:

   - Bert Bos
   - David Carlisle
   - Christopher Comninel
   - Sam Dooley
   - David Farmer
   - Deyan Ginev
   - Patrick Ion
   - Brian Kardell
   - Paul Libbrecht
   - Louis Maher
   - Bruce Miller
   - Steve Noble
   - Murray Sargent
   - Cary Supalo
   - Neil Soiffer
   - Moritz Schubotz
   - Stephen Watt
   - Laurence Zaysser

Regrets:

   - Stephen Watt

Announcements/updatesprogress moving repos

LM: The tpac meeting will be virtual from October 18 through October 29,
2021.

NS: for email subscribe to www-math@W3.org

NS: I will send the announcement to member-math old CG list to subscribe
(including how to subscribe)

BB: there is money available to increase diversity for people to attend
TPAC. Perhaps you can get hardware and maybe office space. BB: there is an
inclusion fund, and another fund for a salary for people who do things for
TPAC. The second fund is the fellows program.

BB: Application currently open for the inclusion fund. More info:
https://www.w3.org/blog/news/archives/9131

NS: we may need to have face-to-face meetings for the MathML WG group
during TPAC.

BK: While we don't want to use IRC for minutes and bots - it still makes
sense to have one for more real time async questions/discussions that might
not qualify for email... So we're re-claiming the old math channel in IRC
(#math). You don't need a IRC client for this, you can do it in your
browser on the w3c itself
https://www.irccloud.com/irc/-irc.w3.org/channel/math or https://irc.w3.org/
channel is "math"
FPWD statusAction Item followup

NS: Got in touch with Wolfram Research to persuade them to join the MathML
WG. Need to have a call with them to address their concerns.

*Action Item* NS: will make a Wiki reminding people about what they
volunteered to do. BK: Mozilla has not joined the WG. They lost funding due
to Covid.

BK: is talking to Google. Igalia is working for them. Apple has not yet
decided to join us. MS works for Microsoft.

LZ: to show that we have widespread adoption, we need two implementations
of something.

NS: MathPlayer is one. he has talked to JAWS. Chrome and Edge may pick up
the Core implementation.

BK: core does not include accessibility, that is the AAM.
https://w3c.github.io/mathml-aam/ (each ML has an AAM and they progress and
measure track advancement a little differently too)

DG: By the end of the WG, MathML Core must ship, for us to claim success.
BK: we must have alignment with browsers.
Follow up on media types issue?

PL: This is about https://github.com/w3c/clipboard-apis/issues/141

PL: No one is talking about clipboard. There is discussion about the use of
UTI. Apple uses UTI.

MS: talked about Microsoft's clipboard format.

NS: should we drop media discussions from MathML 4?

From Laurence ZAYSSER to Everyone: What about the mimetype
application/mathml+xml ?

PL: we should contact the clipboard API developers and JavaScript.

NS: what do we do about people who are not responding. BK and PL: we keep
trying to work with them.

PL: word uses UTI names on Mac. Javascript side of things:
https://w3c.github.io/clipboard-apis/index.html

Issue filed where https://github.com/w3c/clipboard-apis/issues/141
Intent discussion -- defaults semantics (only got started on this last
meeting):

NS: should we invent a new language around intent? Should we be using CSS?
We should talk about defaults. We will need to talk with the accessibility
group.

DG: has implemented three prototypes for intent since joining the CG in
2020. 1) pmml+intent attribute, 2) pmml+cmml parallel trees and recently 3)
pmml+ARIA attributes.

   - 1) and 2) are located at
   https://dginev.github.io/tiny-mathml-a11y-demo/
   - 3) is at https://hackmd.io/@dginev/SkBHsZTiO
   - The ARIA approach is voiced successfully already today, directly using
   Chrome with Screen Reader.

NS: The aria-label approach does not directly work with Braille. DG: A good
opportunity to liaison with the ARIA group and reach a bilateral agreement
on how to use ARIA on MathML, and how AT tools can interpret it.

LZ: reminded people about using altext.
DG: one drawback is that it's not possible to do navigation with altext,
one always needs to narrate the full formula.

LZ: we will need authoring tools to generate what ever system we promote.

DC: we have discussed this for 18 months without a decision. We need a
decision and we should implement that decision.

NS: Action item: next meeting DG will give a demo of his ARIA prototype
implementation.

NS: An issue with using ARIA labels is that it overrides braille, so you
need to include the proper braille code also. ARIA just added a way to
specify braille for a label. It is discouraged to do so in general, but
because math has its own braille code (actually many math codes), it makes
sense for math.

LZ: ARIA can be dedicated to AT. altext can be used for AT and visual
displays.

NS: Meet next on July 15.