Minutes: MathML Full meeting, 7 Sept, 2023

   - Neil Soiffer
   - Louis Maher
   - Sam Dooley
   - Cary Supalo
   - Moritz Schubotz
   - Bruce Miller
   - Paul Libbrecht
   - David Farmer
   - Deyan Ginev

<https://sandbox.cryptpad.info/code/inner.html?ver=5.4.1#cp-md-0-regrets>
Regrets

   - David Carlisle

<https://sandbox.cryptpad.info/code/inner.html?ver=5.4.1#cp-md-0-agenda>
Agenda
<https://sandbox.cryptpad.info/code/inner.html?ver=5.4.1#cp-md-0-1-announcements-updates-progress-reports>1.
Announcements/Updates/Progress reports

NS: At the ARIA meeting, there will be a discussion on how publishers do
not wish to use MathML for inline mathematics. Neil suggests we attend that
meeting.

ARIA discussion:
https://github.com/w3c/publ-a11y/wiki/Accessible-inline-math

NS: The meeting will be during the night Pacific Standard Time.

NS: The TPAC MathML meeting will take place on Thursday, September 14,
2023, at 8 AM to 9:30 AM PST.
<https://sandbox.cryptpad.info/code/inner.html?ver=5.4.1#cp-md-0-2-tpac->2.
TPAC:
<https://sandbox.cryptpad.info/code/inner.html?ver=5.4.1#cp-md-0-a-register-at>a)
Register at

(https://www.w3.org/2023/09/TPAC/registration.html) (there is a box you can
click that says you want a fee waiver -- it is automatic)

PL: TPAC registration will reopen on Monday, September 11, 2023.
<https://sandbox.cryptpad.info/code/inner.html?ver=5.4.1#cp-md-0-b-agenda-for-tpac-still-waiting-for-more-ideas>b)
Agenda for TPAC -- still waiting for more ideas

From our 9/31/2023 intent meeting minutes:

[ NS: Our TPAC meeting is on Thursday, September 14, 2023, for 1.5 hours.
We should do something different from our usual meeting schedule to
interest people who have not been to our meeting previously.

NS is looking for agenda topics.

MuS said he has made a lot of progress on his Unicode to MathML converter.
He would like to demonstrate his application.

MuS said his application is open source and is available for people to use.
It is on GitHub and can get intent into MathML.

NS: It might be good to have a roundup of new MathML generators.

PL: If someone from Apple comes along, then it would be worth clarifying
the discussion about the clipboard Uniform Type Identifier specified in
MathML3 which, apparently, may be a problem with the current Apple
guidelines: public.mathml, public.mathml.presentation and
public.mathml.content.

NS: I have some things we could discuss during the TPAC meeting.

NS: Do we want to plan? We could layout a timeline for next year.

NS: There has been a question about MathML 5 and content MathML.

NS: A part of our new charter is to start thinking about a roadmap for
MathML 5.

NS: We would like to know peoples' implementation plans. At some point we
are going to CR, and we need to collect implementations. We need to have
ways of testing things.

NS: We have been asked how content MathML will relate to intent; therefore,
discussing content MathML is not out of scope.

DM: There is lots of confusion about intent. People project their
understanding of content MathML into intent. We should clarify how they are
different. ]

PL: We can take more time to discuss the major issues such as the topics we
are considering today.

*ACTION* NS will put up a TPAC agenda.
3. Progress/discussion on core concepts/
<https://w3c.github.io/mathml-docs/intent-core-concepts/> core properties
<https://w3c.github.io/mathml-docs/intent-core-properties/> lists

NS: filled out more items on this list.
a) core concepts of one-letter-symbols (472)
<https://github.com/w3c/mathml/issues/472>

DG: gave two reasons for including a letter in the core list. First, if
it's usually spoken as its underlying concept, meaning you don't say the
letter, but you say what it means. He gave the example of when "P" is used
to mean probability, then that is a good reason for inclusion in core.

NS: The second reason to include things in core if they appear in
pre-university STEM education.

DG: There is uncertainty on the border between pre and post university
education. Are we trying to make core as small, or as big, as possible?

NS: Another criteria for inclusion is does the item fit in a group of
things we are planning to consider?

NS: In physics, if we include amperes, are we considering some units and
leaving out others?

DF: The names can apply in other situations. The empty set symbol, {}, can
be used if it is included in core. If it is in core, we have a concept for
it, and it is easy to find.

NS: Letters are spoken as letters. If they have another use, then they need
an intent, and they need to be in core.

NS: Blackboard Bold Z can be used to stand for the set of integers.

NS: said that while "e" and "pi" have special meanings, they should not be
in core because they always have the same meaning.

PL: {} could also mean the empty list.

NS: If you call it something else, give it an intent.

PL: Are we declining terseness?

NS: has added both terse and verbose speech to his core list.

NS: He is not sure we are supposed to tell people how they should speak.

NS: Proposed resolution: given a set of notations with “analogous
meanings”, if there is a significant pronunciation of a notation that is
inappropriate for another, then we should separate the intents of these
notations.

*CONSENSUS:* PL: We agree on the following criteria, things go in core if:

   - if they appear in pre-university STEM, in a textbook published after
   year 2000, and/or
   - the letter has a different pronunciation when read aloud in a formula
   of the simple letter pronunciation.

DG: Sometimes letters work well as properties.

NS: When we talk of the pre-university being the core boundary, we are
limiting ourselves to textbooks.

NS gave his six grade children books which taught advanced concepts like
infinity.
b) derivatives -- differing notations with similar semantics (473)
<https://github.com/w3c/mathml/issues/473>

NS: core concepts: how should the multiple forms of differentiation be
handled?

There are four common notations for differentiation:

   1. Leibniz: Dy/dx
   2. Lagrange: F prime(x)
   3. Euler: DF(x) and
   4. Newton: Y with dot above.

NS: People pointed out there's yet another way to show differentiation, and
that is where you specify as a subscript the variable differentiation to
the Euler notation.

NS: They are really the same concept with the same verbose definition.

NS: What is the concept name? Is there only one concept name? What about
higher order derivatives?

DC, in the issue comments, came out on the side that there is only the
first derivative.

DG: Look at division. He gave five examples. The example using a colon did
not mean ratio, it is the eastern way to show division.

DG: Inventing an intent concept list of "divide-slash", "divide",
"divide-fraction", "divide-colon" and "divide-long" seems counterproductive
here (reason: it feels ad-hoc, while mixing concepts with notations). At
least it is a new kind of convention that we haven't agreed on yet. it
feels ad-hoc, while mixing concepts with notations). At least it is a new
kind of convention that we haven't agreed on yet. What seems mildly more
practical is to anchor the standard concept "divide", and have AT be able
to consult the presentation tree, to determine which notation was used.

DG: is concerned about the predictability of the list. It is easy to drift
into chaos.

DG: For these five examples, the shared operation is division. Do we want
to compress down to one word (division) or give more information. Perhaps
change the intent naming scheme?

BM: For the derivatives: it is tempting to go down to a single word, else
you have so many words.

NS: His worry is that they all represent the same idea so let us have one
name for the concepts. This pushes towards the semantics and not the actual
speech. We need to hear 1 / 2 or 1 divided two and long division is spoken
differently.

NS: We speak and write different types of derivatives, and the blind want
to hear the differences.

NS: Just because they have the same semantics, they should not have the
same concept names.

DG: We are discussing intent design. We need to say things the way they are
heard. We are disambiguating things. How something is written can be
relevant to the way it is spoken.

BM: We use Newton's ideas and Leibniz's notations.

From Moritz Schubotz to Everyone: Dot is often time derivative which is
quite handy to differentiate from other derivatives ''

BM: Those not informed in this field might need a catalog to see how things
are spoken.

PL: The blind want the accurate speech. Time derivatives might not be
wanted if you have prime.

NS: DG suggested we can use hyphenated names or properties to affect the
different ways of speaking things.

NS: For the spec, we want a naming standard that has rationality to it.

DG said that DC approved of the property's method of differentiating speech.

You need to come up with a plan on how to separate the notational detail
from the concept detail rather than create these hybrid concepts that have
both things included in them.

NS: Would you speak the colon and slash?

PL: Three fifths would be a fraction and never a division.

DG: You can say one over two if it is a colon or a division.

NS: It is difficult to have a hard and fast rule about how things should be
spoken.

NS: There are different ways to write derivatives which are spoken
differently from time to time.

NS: Make sure that core is not special or different than open in a
significant way. It's different in that AT needs to know about it and may
have specialized rules for reading things as in the division case with a
half versus one over 2.

DG: Keep core and open consistent. If it is in core, it gets special
treatment. We must have tolerance for special treatment in core.

Paul's proposal: given a set of “analogous meanings”, if there is a
significant pronunciation of a notation that is inappropriate for another,
then we should separate the intents of these notations.

NS: may add this discussion to the TPAC agenda.

DG: We should think about how to make the naming procedure not messy. We
want the procedure to be understandable and intuitive.

NS: We want predictability.

DG: We want the same naming scheme in core and in open.

DG: In core, you have the luxury of special treatment.

NS: There is a little bit of a difference between the fractions example and
the derivative example in that most of those fraction notations not all but
most of them involved a symbol, a Unicode symbol, which has kind of a name
in a way of being spoken, whereas the derivatives are really notations that
are built up and not a single symbol. So, they are two different things and
may have different solutions.