Minutes: MathML core meeting, 3 Nov, 2022

 Attendees:

   - Neil Soiffer
   - Louis Maher
   - David Carlisle
   - Dennis Mueller
   - Sam Dooley
   - Johannes Stegmüller
   - Bert Bos
   - Patrick Ion
   - Bruce Miller
   - David Farmer
   - Deyan Ginev
   - Paul Libbrecht
   - Murray Sargent

<https://sandbox.cryptpad.info/code/inner.html?ver=5.2.0-rc2#cp-md-0-regrets>
Regrets

Steve Nobel
<https://sandbox.cryptpad.info/code/inner.html?ver=5.2.0-rc2#cp-md-0-agenda>
Agenda
<https://sandbox.cryptpad.info/code/inner.html?ver=5.2.0-rc2#cp-md-0-1-announcements-updates-progress-reports>1.
Announcements/Updates/Progress reports

NS: MathML is turned on in Chrome Canary. Chrome (version 109), with
MathML, may be shipped on January 9, 2023.

NS: is giving a talk at an accessibility meeting. He wanted to list some
software that might change intent. PreTeXt, LaTeXML.

DC: is working on some code that might do this (
https://w3c.github.io/mathml-docs/ctopintent/)

JS: We're currently working on a software for media Wiki. It should produce
MathML and intent attributes. It is in the math extension and is written in
PHP.

DM: is working on STeX to produce intent.

NS: We should have a meeting to let people report on their progress, and so
people can indicate problems in the spec.

DM: Has a student working on a screen reader plugin to read intent.

DG: Some of DG's time will be spent investigating a MathJax plugin to
generate intent.

JS: Is working on a converter from LaTeX to MathML. It uses a JSON file.

NS: Please add the binomial to this list of math expressions that should be
mapped to MathML.

JS: He would like suggestions for what should be added.

From Deyan Ginev to Everyone:
https://www.onemathematicalcat.org/MathJaxDocumentation/TeXSyntax.htm List
of TeX commands available in MathJax.
<https://sandbox.cryptpad.info/code/inner.html?ver=5.2.0-rc2#cp-md-0-2-continue-discussion-on-39-intent-39-name-list-registry-spreadsheet->2.
Continue discussion on 'intent' name list/registry/spreadsheet/...

NS: In DG's table, NS said that there were only two fields people agreed
upon: the name and the MathML that should come out of it.

LM: suggested talking about each column heading and deciding which ones
people wanted to keep.

NS: proposed keeping the subject area column.

Ns: How would the subject area category help someone who is looking for a
specific intent?

DF: It would help distinguish between two things that look the same.

NS: How could you have a default intent if you did not have a subject area.

DG: Supported a Type column.

DC: Subject areas would be useful to define defaults.

NS: says equal signs has a default

NS: Make a separate table of defaults.

NS: Consider a single prime could be foot, a second or a derivative, or
just a marker

BM: Think of different tasks people might want to do. If you are given a
MathML and subject area, you will look for MathML patterns.

NS: Who is going to use this table: 1. Implementers, who need to generate
intent for a construct. 2. Remediators, who have MathML that is spoken
poorly, how can they make it speak more correctly.

NS: AT implementers are not going to be primary users. AT will speak intent
as it is. AT will use intent for things that are not obvious. A fraction
has many ways to say it.

NS: Unicode operators are self-voicing for the most part. Only things like
fractions and powers which have different ways of being pronounced will
need help from intent.

BM: The number of AT intent users is small.

From MurryS to Everyone:
https://www.onemathematicalcat.org/MathJaxDocumentation/TeXSyntax.html
list: (LM: This web address is not correct. I may have made an error while
copying it from the chat.)

BM: People with different tasks to accomplish may want different columns in
different order. It would be desirable to decide which columns you wanted
and assemble a new table sorted on those columns.

DC: What are we going to use this for?

PI: Greater than can be replaced by exceeds. There are many ways to say
some things.

NS: The simpler the table is, the simpler it is to use and add to. NS would
like to reduce the fields in the table.

PL: Some things are so simple they should not have anything in the intent
field.

MUS: The simpler the better.

DG: 'greater than zero' could be said as 'positive'.

NS: Z squared could mean two dimensions of Z, i.e., the Cartesian product.

DF: limit as x goes to one from the right.

DF: M is positive, but M could be a matrix which means it is positive
definite.

PI: Discussed 'blackboard bold Z sub /3' meaning the integers mod 3, or the
finite field of order 3. (Or use 12 instead of 3 for clocks, admittedly not
a field then.)

PL: You need intent to know the meaning of positive.

BM: We are slipping into a territory where we have different assumptions on
what we are trying to accomplish. Are we doing an alternative to aria with
more math awareness? Different people are looking at it in different ways.

DC: We have had this question for three years. Core has around 200 lines.
If we want to make a smaller table, we should just write down a table so
that we can make progress. If we want to make it smaller, put down a list
of 50 rows and two columns. We must do something to move on. Just make a
smaller list.

NS: For the core list, have an open intent which could give guidance on
what things should be called.

NS: There have been no studies of syntactic versus semantic reading.

NS: Math speak will let you know what is there exactly.

DG: If you want a small table, start fresh from the core list, and just do
it. The open list is not directly connected with the group. Everyone can
decide what columns they want. This small list will cover 50 to 200 items
and does not cover K-14. The goal of intent is to facilitate good speech
generation.

BM: Before we throw DG's table out, decide who are the users. Decide who
needs this table for what purposes? Then everything else is optional.

NS: We have a small list of two columns saved: the name and the MathML
produced.

BM: For an annotator, things may look different on different browsers and
may be spoken differently. We would like core to keep pronunciation the
same across browsers.

DC: asked BB what DC should do to help the specification release.

NS: Needs to get out the horizontal review announcement before CR.

NS: asked BB: What else needs to be done?

DC: Unicode 15 came out in September. we should account for this.

DC: asked BB to email him if DC needs to do something.

NS: What do people want to talk about? What is the point of the table and
intent?

DC: Make a small table to start discussions. We must start somewhere.

NS: Who wants to make the small table?

NS: We will not discard parts of DG's work. We will copy the file and make
new tables to see what would be useful.

DC and NS will work on making the small list.