Minutes: MathML Full meeting 11 Dec, 2025

 Attendees:

   - Neil Soiffer
   - Louis Maher
   - Paul Libbrecht
   - David Carlisle
   - Moritz Schubotz
   - Deyan Ginev
   - Bert Bos
   - Bruce Miller
   - Murray Sargent

<https://cryptpad.fr/#cp-md-0-regrets>Regrets
<https://cryptpad.fr/#cp-md-0-action-items>Action Items
<https://cryptpad.fr/#cp-md-0-2-any-issues-and-or-pending-work-that-needs-to-be-done-before-a-cr-vote->2.
Any issues and/or pending work that needs to be done before a CR vote?

*ACTION:* DC has work pending.

*ACTION:* Discuss the accent characters in the December 15 2025 core
meeting issue 554 Intent of MathML-Core B.3 is unclear
<https://github.com/w3c/mathml/issues/554>.

*ACTION:* MoS: As we move to CR, MoS was worried about the cross-links
between sections. He wants more proofreading to be done keeping in mind how
one section relates to other sections. MoS may open an issue or suggest a
PR on this.
<https://cryptpad.fr/#cp-md-0-agenda>Agenda
<https://cryptpad.fr/#cp-md-0-1-announcements-updates-progress-reports>1.
Announcements/Updates/Progress reports
<https://cryptpad.fr/#cp-md-0-firefox-pdf-progress>Firefox PDF progress

NS: The Firefox implementers have fixed most of the pdf related math.

DC: We are getting good response on fixing Firefox bugs.

DC believes the Firefox pdf math work should be released in February.
<https://cryptpad.fr/#cp-md-0-spec-updates>Spec updates

NS: We have a number of spec updates.

DC has approved NS' naming of the decimalpoint literal's value name to be
decimalcharacter. just noting a matching change in the MathML-schema repo
is needed.

In DG's work allowing case to be ignored there is a line mentioning this in
every instance of where it is used. We had a discussion on whether we could
say this just once. The decision was to leave it in place because core
mentions it every time it is used.

From Moritz Schubotz to everyone: I think the advantage is that you don’t
need to read the whole document to see the mixed case comment.

NS: mstyle has one difference between core and full and that is that mstyle
does not have default element attribute values like linethickness for mrow
in core. That is not one of the global attributes.

NS: Should every element be able to set the line thickness, or just keep it
on mstyle?

Consensus to leave it on mstyle.
<https://cryptpad.fr/#cp-md-1-2-any-issues-and-or-pending-work-that-needs-to-be-done-before-a-cr-vote->2.
Any issues and/or pending work that needs to be done before a CR vote?

*ACTION:* DC has work pending.

DC has malign/maligngroup work to do. After that the others are editorial.

NS does not know what to do about the accent characters.

DC thought that the MathML core spec appendix B.3 tried to answer this
question: Intent of MathML-Core B.3 is unclear, issue554
<https://github.com/w3c/mathml/issues/554>.

*ACTION:* Discuss the accent characters in the December 15 2025 core
meeting issue 554: Intent of MathML-Core B.3 is unclear
<https://github.com/w3c/mathml/issues/554>.

PL wanted to have discussions with companies asking them to conform with
font standards, especially stretching.

MuS has been recalled by Microsoft to get math accessibility into Microsoft
products.

MuS noted that <A> using href does not work in MathJAX. How do we tell
MathJAX to fix it?

NS: I can write to them, but first we have to get it into core. Do the
browsers support it?

MuS: It works in Chromium.

*ACTION:* MoS: As we move to CR, MoS was worried about the cross-links
between sections. He wants more proofreading to be done keeping in mind how
one section relates to other sections. MoS may open an issue or suggest a
PR on this.

NS discussed the need for the spec to be reviewed by other W3C groups like
security, Accessibility, internationalization, and others. These groups
would be given about a work month to review the spec.

NS: It is not a requirement that the groups approve it. It is a matter of
politeness to ask groups to review our work.

BB thought that a month was a reasonable time for the other groups to
review our work.

NS: This gives us extra time to do proofreading.

NS: The W3C information release process has been simplified.

NS hoped that these reviews could be finished by February one.

PL thought that the international group would worry about the format of our
messages generated by intents.

NS believes that all the decisions are already made and they just need to
be written up.

NS hopes that the MathML 4 process can be wrapped up by October 2026.

NS wants to get to CR by February 1 2026.
<https://cryptpad.fr/#cp-md-0-zoom-meeting-summary>Zoom Meeting Summary
<https://cryptpad.fr/#cp-md-0-summary>Summary

The team discussed various technical implementations, including PDF
rendering support in Firefox and schema validation issues. They reviewed
MathML4 specifications and discussed ongoing work on font rendering, accent
characters, and browser compatibility, with plans to move the standard to
Candidate Recommendation status by February 1st. The conversation ended
with discussions about attribute handling in core specifications and the
need for proofreading and cross-linking between document sections.
<https://cryptpad.fr/#cp-md-0-pdf-rendering-and-math-issues>PDF Rendering
and Math Issues

The team discussed progress on PDF rendering support in Firefox, with David
reporting that while the implementation is working well, there are still
some math-related issues that need to be addressed. David noted that the
fix should be included in Firefox's beta release by February, with a 4-week
cycle rather than the previously expected 6 weeks. The team also touched on
Microsoft's AI initiatives and their work on improving Word's mathematical
capabilities.
<https://cryptpad.fr/#cp-md-0-schema-validation-discussion>Schema
Validation Discussion

David and Neils discussed schema validation issues, particularly regarding
mixed-case patterns and regex support. They agreed that while the current
schema implementation is "ugly" and limits editor functionality, they
should proceed with the specification PR without being overly concerned
about schema technicalities, as these can be adjusted later. Paul noted
that XML editors like IntelliJ and Visual Studio typically attempt to fetch
schemas for validation, though the impact on real-time editing is minimal.
<https://cryptpad.fr/#cp-md-0-case-sensitivity-and-m-style-attributes>Case
Sensitivity and mstyle Attributes

The team discussed handling case-insensitive matching in specifications,
with David suggesting adding a link to an explanation rather than repeating
specifications. They agreed to leave the current approach as-is for now,
pending feedback. Neil presented a difference between Core and full
versions of mstyle attributes, particularly regarding line thickness
settings, and sought input on whether these should be universally
applicable or limited to mstyle elements.
<https://cryptpad.fr/#cp-md-0-mathml-attributes-discussion>MathML
Attributes Discussion

David and Neil discussed the attributes in core and their placement, noting
that some attributes are specific to MathML presentation elements. They
agreed that if changes were made to core attributes, such as line
thickness, they would need to be updated across the board. Neil confirmed
that the script and line thickness attributes can still be set on an
mstyle, maintaining its specialist status. The conversation ended with no
additional announcements or progress reports.
<https://cryptpad.fr/#cp-md-0-mathml4-progress-and-cr-planning>MathML4
Progress and CR Planning

The meeting focused on the status and next steps for MathML4, including
discussions on font rendering issues, accent characters, and browser
compatibility. David highlighted ongoing work on font changes and rendering
inconsistencies, while Paul suggested creating a test suite for rendering
across different browsers. Murray mentioned progress on integrating MathML4
into Microsoft products and proposed adding the A tag for hrefs, which
David agreed to address. Moritz emphasized the need for proofreading and
cross-linking between sections of the document. The group discussed the
timeline for moving MathML4 to Candidate Recommendation (CR) status, aiming
for February 1st, and the need for approval from various review groups.
Bert confirmed that all review groups must have no open issues before CR
can be achieved.