Minutes: MathML intent meeting 4 Nov, 2021

 Attendees:

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

Regrets:

   - Daniel O'Mahony
   - Moritz Schubotz

Announcements / UpdatesPresentations by Bruce and Sam on their intent ideas

Bruce's material:

https://mathml-refresh.github.io/discussion-papers/semantics-mini

Deyan's demo page (with some options selected) mostly reflects Bruce's
proposal: https://dginev.github.io/tiny-mathml-a11y-demo/

Sam's material:
https://samdooley.github.io/mathml-docs/intent2cmml/intent.html

BM: reviewed his paper.

LM: Said that Sam used the "apply" attribute where Bruce did not.

BM: Sam is thinking about generating content MathML. Bruce says "apply" is
implied in his 's method.

PL: In Deyan's demo, instead of apply, he got times signs.

DG: LaTeX ML assumes that things it can't understand are interpreted as
multiplications; therefore, PL saw many times signs when he ran the demo.

BM: His proposal meets the technical requirements for disambiguation and
accessibility. He is not sure if it fits the market's requirements--will
people implement his method?

BM: Discussed literals and selectors.

BM: A literal is a keyword to be gotten from a dictionary, or it's a
literal phrase to be spoken as is.

BM: Defaulting is a separate discussion. Hopefully, it uses the same
language as intent.

BM: The defaulting process would not apply to things that had explicit
intents.

BM: Intent must have something that says that defaulting does not apply.

NS: How do you know when defaulting starts?

BM: If intent points to a child, then the child is invisible to the
defaulting system.

BM: When choosing names, one should consider names that people can
remember. Do you put in helping words such as "of" as in "the derivative
of"?

BM: Showed his method of handling the case where the integration variable
was in a numerator.

SD: Started reviewing his paper.

SD: said he had concentrated on the issue of disambiguation for
presentation MathML.

SD: Said that he and BM agreed on many things.

SD: Showed a prototype translator which turns intent MathML into content
MathML. This translator is not complete, and SD did not share its link.

SD: showed a demonstration of how this process worked to convert a
presentation MathML example of the quadratic equation into content MathML.

SD: showed several examples that were in his paper.

SD: showed his integral example where the integration variable was in the
numerator.

SD: We have not yet had a dictionary discussion. Also, we have not had a
discussion on how we produce natural language.