Minutes: MathML Full meeting, 23 April, 2026

 Attendees:

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

<https://cryptpad.fr/#cp-md-0-regrets>Regrets

   - Paul Libbrecht

<https://cryptpad.fr/#cp-md-0-action-items>Action Items
<https://cryptpad.fr/#cp-md-0-1-announcements-updates-progress-reports>1.
Announcements/Updates/Progress reports

Horizontal review status for CR

*ACTION:* BB will check the status of the CR process.
<https://cryptpad.fr/#cp-md-0-2-next-steps->2. Next steps?

*ACTION:* NS should talk to the WCAG group about considering intent.
<https://cryptpad.fr/#cp-md-0-3-issues-to-review->3. Issues to review:
<https://cryptpad.fr/#cp-md-0-a-close-559->A close 559?

processContents="skip" for other attributes of common MathML elements
<https://github.com/w3c/mathml/issues/559> (related: 498 xsd)
<https://github.com/w3c/mathml/issues/498>

*ACTION:* DC: will make a MathML4 schema that works. Leve
processContents="skip"
for other attributes of common MathML elements issue 559
<https://github.com/w3c/mathml/issues/559> Open.
<https://cryptpad.fr/#cp-md-0-b-where-are-we-on-522-a-href-https-github-com-w3c-mathml-issues-522-nested-math-in-token-elements-a->b.
where are we on 522: nested math in token elements
<https://github.com/w3c/mathml/issues/522>

*ACTION:* DC: change the MathML4 spec text to allow MathML in tokens and
close nested math in token elements issue 522
<https://github.com/w3c/mathml/issues/522>.
<https://cryptpad.fr/#cp-md-0-c-where-are-we-on-560-a-href-https-github-com-w3c-mathml-issues-560-mathvariant-attribute-enumeration-is-missing-chancery-and-roundhand-a->c.
where are we on 560: mathvariant attribute enumeration is missing chancery
and roundhand <https://github.com/w3c/mathml/issues/560>

*ACTION:* DC wrote: decided to close mathvariant attribute enumeration is
missing chancery and roundhand issue 560
<https://github.com/w3c/mathml/issues/560> as mathvariant is not in core
(amongst other issues. Will add a note that variant selector FE00 FE01 may
be used to disambiguate script. In the spec, it should be mentioned that
we're not adding these styles of fonts.
<https://cryptpad.fr/#cp-md-0-4-thoughts-on-a-href-https-github-com-w3c-mathml-core-issues-320-mathml-core-issue-320-browser-status-a->4.
Thoughts on MathML Core issue 320: Browser Status?
<https://github.com/w3c/mathml-core/issues/320>

*ACTION:* NS will add this list (https://temml.org/docs/en/mathml-status)
to the Monday April 27) core meeting agenda.
<https://cryptpad.fr/#cp-md-0-agenda>Agenda
<https://cryptpad.fr/#cp-md-1-1-announcements-updates-progress-reports>1.
Announcements/Updates/Progress reports

Charter status

NS: Has there been any progress on the charter or CR?

BB: The charter has been reviewed by two of the five horizontal groups.
Thus far, there have not been any issues.

Horizontal review status for CR

BB: The charter and CR processes are independent.

*ACTION:* BB will check the status of the CR process.

Issues for core meeting?

Thoughts on MathML Core issue 320: Browser Status?
<https://github.com/w3c/mathml-core/issues/320>

NS: This year's TPAC meeting takes place from October 26 through 30 2026,
in Dublin. The 2027 TPAC meeting takes place from December 6 – 10, 2027, in
Panama.

DC: The PDF Association, in collaboration with the LaTeX team on Accessible
PDF math documents, has just published the New Best Practice Guide for
Mathematical Content in PDF (
https://pdfa.org/resource/best-practice-guide-math-in-pdf/). MathML plays a
prominent role in tagged PDF for such documents.

NS said that he recently published a paper that had to use the Springer
style guide. That guide caused the paper to be inaccessible.

DC said that he knew how to make the Springer publishing requirements
accessible.
<https://cryptpad.fr/#cp-md-1-2-next-steps->2. Next steps?

NS: What should we be doing until the CR gets approved?

NS: There are plenty of issues which need work.

NS: The core concept lists need work.

NS: MoS and DG are working on a project that needs to be completed.

NS: Polyfills need work.

DC: There are several issues that need a spec update before they can be
closed.

DG: We need to write tests. Do we need to test intents?

NS: With the web platform tests the only thing you can say is yes, it
accepts intents.

NS: We could look outside of web platform tests.

NS: A group within the Accessibility Platform Architecture Group (APA) is
working on testing how screen readers react to MathML.

DG: We had requested and received an audit from Cornell University for
archives base web pages for accessibility, for the deadline that is now, at
least for us, is extended by one year. We had to comply with the WCAG 2.2
guidelines.

From Moritz Schubotz to everyone: For tests, I think it would be nice to
align the tests to corresponding TeX input.
https://github.com/wikimedia/mediawiki-extensions-Math/blob/master/doc/main.pdf
It occurs to me that the user consider TeX as the gold standard how things
are supposed to look (that also seems to be aligned with
https://temml.org/docs/en/mathml-status where the exact Tex rendering is
realized with a heavy css massaging)

NS: The WCAG guideline said that if math was in your document, then it
should be written with MathML. We should get them to consider intent.

*ACTION:* NS should talk to the WCAG group about considering intent.

DG: The thing I like about them is they have validation tools which are
tied into tests.

DC: It's hard to test intent because intent generates speech.

NS: Maybe the ARIA Group is doing testing with screen readers to make sure
like, if we set ARIA label, are they actually using ARIA label?

DC: You mean the testing by sitting there and listening to it, or do they
have some tests acting automatically.

NS: They may have come up with a method to test screen readers
automatically.

From Deyan Ginev to everyone: I wanted to raise the topic. I am sure we
need multiple meetings to solve the testing topic
<https://cryptpad.fr/#cp-md-1-3-issues-to-review->3. Issues to review:a.
close 559?

processContents="skip" for other attributes of common MathML elements
<https://github.com/w3c/mathml/issues/559> (related: 498 xsd)
<https://github.com/w3c/mathml/issues/498>

DC: Too late for Mathml3. not done for MathML 4.

From Moritz Schubotz to everyone: @Deyan Ginev I think it’s on the list
now, isn’t it?

DC: So I might just hand-write the XSD version of the schema for MathML 4.

*ACTION:* DC: will make a MathML4 schema that works. Leve
processContents="skip"
for other attributes of common mathml elements issue 559
<https://github.com/w3c/mathml/issues/559> Open.
<https://cryptpad.fr/#cp-md-1-b-where-are-we-on-522-a-href-https-github-com-w3c-mathml-issues-522-nested-math-in-token-elements-a->b.
where are we on 522: nested math in token elements
<https://github.com/w3c/mathml/issues/522>

NS: This is the one where we have MathML nested in MathML.

DC: The text needs changing.

*ACTION:* DC: change the MathML4 spec text to allow MathML in tokens and
close nested math in token elements issue 522
<https://github.com/w3c/mathml/issues/522>.
<https://cryptpad.fr/#cp-md-1-c-where-are-we-on-560-a-href-https-github-com-w3c-mathml-issues-560-mathvariant-attribute-enumeration-is-missing-chancery-and-roundhand-a->c.
where are we on 560: mathvariant attribute enumeration is missing chancery
and roundhand <https://github.com/w3c/mathml/issues/560>

NS: There is no need to change the schema because it's already allowed.

NS: I do want to remind everyone that the Mathvariant is sort of almost an
obsolete thing, in that, people just can put the Unicode equivalent.

NS: Core does not support this mathvariant.

DC: We could just not do it, and just say, mathvariants are deprecated
anyway.

MuS: In the spec, it should be mentioned that we're not adding it.

*ACTION :* DC wrote: decided to close mathvariant attribute enumeration is
missing chancery and roundhand issue 560
<https://github.com/w3c/mathml/issues/560> as mathvariant is not in core
(amongst other issues. Will add a note that variant selector FE00 FE01 may
be used to disambiguate script. In the spec, it should be mentioned that
we're not adding these styles of fonts.
<https://cryptpad.fr/#cp-md-1-4-thoughts-on-a-href-https-github-com-w3c-mathml-core-issues-320-mathml-core-issue-320-browser-status-a->4.
Thoughts on MathML Core issue 320: Browser Status?
<https://github.com/w3c/mathml-core/issues/320>

NS: It is a very nice list about all the things that are wrong with
browsers. The list was started by MuS.

NS: It would be good to talk about this at the next core meeting.

DC was interested in item 4: Operator spacing.

DG: Process suggestion. I like the list a lot. It would be nice to give
somebody very senior and well knowledgeable on the topics, the task, to
check if the ones that don't have an issue should have a new issue. And not
only for the S status, but for all of them. It would be great if we track
all of them, and then we have confidence everything is in the issues. And
then people that work on improving browsers can improve the browsers. Maybe
the group can do it in a meeting.

DC: This list is obviously on the author's site. We could add another
column which had an issue link or something. I mean, how do we want to
proceed? This file has issue links, but we cannot change this file because
it belongs to the author.

DG: We can give him a comment in the GitHub issue he opened, and then he
can update his sheet.

DC: Can we take this file and copy it into our own GitHub and check it
ourselves?

From Moritz Schubotz to everyone: The file seems very much aligned with the
issues I ran into so that seems to be an excellent start

DC: If we copy the file, there will be two copies of the file and that can
cause problems.

NS: We will discuss this file in the Monday core meeting.

NS: I think probably we can focus on a few of the issues that are listed
here as spec issues.

*ACTION:* NS will add this list (https://temml.org/docs/en/mathml-status)
to the Monday April 27) core meeting agenda.
<https://cryptpad.fr/#cp-md-0-zoom-meeting-intent-summary-4-23-2026>Zoom
Meeting Intent Summary 4/23/2026 <https://cryptpad.fr/#cp-md-0-summary>
Summary

The Math Working Group held a meeting to discuss progress updates and
review open issues. Neil led the discussion on charter and CR status,
noting that the charter had been reviewed by two of five horizontal groups
while waiting for tech, security, and privacy reviews. Neil mentioned the
upcoming TPAC events in Dublin and Panama, and reviewed browser rendering
issues that needed to be addressed in the specification. Key technical
discussions centered around MathML schema updates, nested MathML elements,
and math variant selectors. The group debated whether to add new math
variant values (Round Hand and Chancery) but ultimately decided to close
the issue due to technical complexities and the fact that the math variant
attribute is deprecated in core. The conversation ended with a review of
browser rendering status issues that would be discussed at an upcoming core
meeting, with Neil planning to add these topics to the agenda.
<https://cryptpad.fr/#cp-md-0-charter-and-process-updates-meeting>Charter
and Process Updates Meeting

The team discussed updates on the charter and CR processes, with Bert
reporting that the charter has been reviewed by two of five horizontal
groups but is still awaiting reviews from tech, security, privacy, I-18N,
and Accessibility groups. Neil announced that TPAC will be held in Dublin
in October and Panama in December. The team also noted an upcoming core
meeting on Monday, with a MathML core issue regarding browser status which
is today's fourth agenda item.
<https://cryptpad.fr/#cp-md-0-pdf-accessibility-projects-discussion>PDF
Accessibility Projects Discussion

The group discussed several ongoing projects and challenges related to PDF
accessibility and document tagging. David announced the publication of a
new PDF Association document on math in PDF, which includes work on MathML
tagging. Neil shared his experience with accessibility issues when
publishing a paper for a conference, noting problems with Springer's style
guide. The group also discussed the US government's extension of
accessibility requirements for some cities but not universities. Neil asked
the group about priorities for their work until CR gets approved,
mentioning areas like core concept lists and polyfills that need attention.
Deyan raised a question about testing intents, suggesting they could look
beyond web platform tests to screen reader handling and testing.
<https://cryptpad.fr/#cp-md-0-mathml-and-wcag-integration-discussion>MathML
and WCAG Integration Discussion

The team discussed the integration of MathML with WCAG 2.2 guidelines,
noting its current inclusion in the techniques section but considering the
potential for further alignment, especially regarding intent expressions.
Neil agreed to initiate a conversation with the WCAG group about
incorporating MathML intent into the guidelines. The group also explored
validation tools and testing approaches for MathML, acknowledging
challenges in browser-based testing and discussing the possibility of
leveraging existing screen reader testing protocols. Next steps include
investigating potential partnerships with the ARIA group for testing and
exploring ways to better integrate MathML into web accessibility standards.
<https://cryptpad.fr/#cp-md-0-mathml4-schema-development-discussion>MathML4
Schema Development Discussion

David and Neil discussed the status of XSD schema for MathML4, with David
explaining that while the current schema is not complete due to issues with
Trang, he plans to create a working MathML4 schema rather than relying on
MathML3. They agreed to allow nested MathML within token elements, with
David committing to update the specification to reflect this decision.
David also noted that the schema already supports this functionality based
on its MathML Core foundation, making the change relatively straightforward.
<https://cryptpad.fr/#cp-md-0-math-font-variant-implementation-discussion>Math
Font Variant Implementation Discussion

The team discussed the implementation of math font variants in Computer
Modern, which now includes three script styles: default, round hand, and
Chancery. They debated whether to add new variant values for round hand and
Chancery to the math variant attribute, but expressed concerns about the
technical complexity and potential issues with Unicode implementation. The
group ultimately considered deprecating the math variant attribute
altogether, given its limited support and the availability of direct
character encoding in modern systems.
<https://cryptpad.fr/#cp-md-0-unicode-fonts-support-discussion>Unicode
Fonts Support Discussion

The team discussed an issue regarding Unicode fonts and character support,
with David explaining that Unicode fonts can support full alphabets while
8-bit fonts are limited to uppercase only. Murray and Neil agreed to close
the issue since additional support wouldn't be implemented in core, and
David committed to adding a note to the specification about variant
selectors FE00 and FE01. The conversation ended with a discussion about
browser rendering issues, with Neil suggesting these topics should be
addressed at the core meeting.
<https://cryptpad.fr/#cp-md-0-mathematical-notation-spacing-rules>Mathematical
Notation Spacing Rules

The team discussed spacing rules in mathematical notation, particularly
focusing on how MathJax and other systems handle spacing around operators
and functions. Neil suggested that if multi-character operators are
involved, they should use thin spacing rather than zero space when adjacent
to other operators. Moritz emphasized the importance of defining clear
rules for translating LaTeX spacing information to MathML, noting that
currently 50% of the master model's size carries this spacing information.
Deyan proposed having a senior team member review the list of issues to
ensure all problems are properly tracked in GitHub issues. Neil agreed to
add this topic to Monday's agenda, with a focus on addressing spec issues
that don't currently have open issues.