Minutes: MathML Full meeting, 8 Jan, 2026

 Attendees:

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

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

We discussed DC's pull request add an element to mathml-code #142 by David
Carlisle <https://github.com/w3c/mathml-core/pull/307>.

*ACTION:* NS will clarify mstyle in full.

*ACTION:* DC: I'll make a comment in issue 555 Add <a> to spec
<https://github.com/w3c/mathml/issues/555> which will add a comment about
attributes.

*ACTION:* DC: will update the new HTML checker
<https://validator.w3.org/nu/about.html#issues>. It is not allowing intent.
Fix this.
<https://cryptpad.fr/#cp-md-0-2-moving-to-cr-status>2. Moving to CR: Status

*ACTION:* DC: To move to CR, DC must work on malign (mostly documentation
changes) and the linking with mrow (which is new).

*ACTION:* For the full committee: pick some sections from the full spec and
see if they make sense, and do not have spelling errors.

*ACTION:* NS will send out the horizontal review request.
<https://cryptpad.fr/#cp-md-0-agenda>Agenda
<https://cryptpad.fr/#cp-md-1-1-announcements-updates-progress-reports>1.
Announcements/Updates/Progress reports

We discussed DC's pull request add a element to mathml-code #142 by
davidcarlisle <https://github.com/w3c/mathml-core/pull/307>.

DC: The whole reason for having A is to make it the same as HTML in SVG.

DC: We say that mrow and <A> are identical because they both have href
already in full.

NS: If we really want to keep people using mrow in full, then it really
should have all the capabilities of an <A>.

From Deyan to everyone: please don't add "type" to mrow It's confusing, but
not a technical issue.

From Moritz Schubotz to everyone: I think it’s just confusing, not ambiguous

MuS: A, in HTML, A can just have some text inside, which is not an MI, it's
not an M text, it's not an MO or an MN or anything like that. Would we be
allowing just ordinary text to be inside the A?

DC: No, because you can already use an A inside an MI, because you get an
HTML A, so that's already allowed.

NS: We probably should note that the contents of the A needs to be valid
MathML, and it can't just be plain text.

NS: So your point is, if we really want to keep people using MROW in full,
then it really should have all the capabilities of an A.

NS: Does anybody feel that mrow in MathML full should not take all the
attributes?

DG: Yes. There is one attribute called type. There's a type attribute in
Content MathML, and there's a type attribute on A, and they're very
different. If we add them to MROW it will be quite confusing.

*ACTION:* NS will clarify mstyle in full.

*ACTION:* DC: I'll make a comment in issue 555 Add <a> to spec
<https://github.com/w3c/mathml/issues/555> which will add a comment about
attributes.

MoS: I had a comment, regarding the discussion with the type that reviewed
on A. I would argue against it. I would say it's not a problem to have the
type argument there.

MoS: He had a meeting with the Wikipedia Foundation, and they said, yeah,
they want to focus in the first quarter of 2026 to get MathML new rendering
deployed, and get the SVG machinery Switched off.

MoS: Some people complained about this loudly. They said the spacing looked
different than it did in TeX and SVG. MuS proposed going with MathJax.

MuS: But there was one… one blocker, and that's, that was a performance
review by the Wikimedia guys, of the MathJax, and it seems they are unhappy
with it because the size of the JavaScript that is loaded with the Mathjax
is too big.

MuS: So, we are looking into that, and if that can be resolved, we can
continue with the plan, and if not, we need to think again what to do.

DG: I made the optimistic decision to test an archive page with the new
validator, the HTML5 validator, and it stopped at 1,000 errors, telling me
a thousand times, this intent attribute is really not allowed on math.

DC: Hopefully, we can just give them the updated HTML core schema.

DC: I mean, the easiest thing is just to take the one they're using and add
the intent argument to it.

DC: It makes no sense to update MathML, and not its checker.

*ACTION:* DC: will update the new HTML checker
<https://validator.w3.org/nu/about.html#issues>. It is not allowing intent.
Fix this.

From Deyan to everyone: the repository is at
https://github.com/validator/validator
<https://cryptpad.fr/#cp-md-1-2-moving-to-cr-status>2. Moving to CR: Status

*ACTION:* DC: To move to CR, DC must work on malign (mostly documentation
changes) and the linking with mrow (which is new).

*ACTION:* For the full committee: pick some sections from the full spec and
see if they make sense, and do not have spelling errors.

*ACTION:* NS will send out the horizontal review request.
<https://cryptpad.fr/#cp-md-0-3-plans-for-2026>3. Plans for 2026

NS: Our charter will expire in the fall. What do we need to meet on this
year?

PL: We need to meet on issues people bring up about the charter.

NS: Perhaps we should discuss MathML 5 and implementations.

MuS: Wanted to add more font attributes. He made an issue for this.

DC: I've got an issue in the TR25 repository, About the math alphabets
being Basically unusable. Maybe we should raise it there.

MS: MathJax did some things that were not great with the new Unicode arrow
characters that we added for chemistry, but it turns out there are 3 more
symbols in chemistry that are about bonds. We should add them to Unicode
also.

DG: I was thinking about outreach. There are two new browser engines coming
out, Ladybird <https://ladybird.org/> and Servo <https://servo.org/>. Servo
is the larger program. Currently Ladybird has as much funding as servo.

DG wants to get their opinion on MathML core and the intent attribute. We
could ask for their opinion on the spec.

DG: These are new browsers from the ground up.

DG: Servo came out of the Firefox space, they did a Rust rewrite to some of
the core, and now they're spun out into the Linux Foundation, and they're
doing the whole thing in Rust.

Copilot: Ladybird is a brand new, fully independent browser engine being
built completely from scratch — no Chromium, no WebKit, no Gecko.

DG: We could ask them what they think of MathML core.

DC: We should be getting core in full.

DC says: we will be pressed to get the specs finished.

PL: We should do more tests.

PL: We want web platform tests and user-oriented tests.

PL: We should continue to have regular meetings to keep us concentrated on
the work.

Ns does not know if we have the need for weekly meetings.

NS: In a standard week, if committee members do not give NS agenda items by
Tuesday evening, then he may cancel the Thursday meeting.

NS: Keep your brains open for a weekly topic, or we may not have a meeting
each week.

DC will concentrate on getting core and full through the process, and not
on getting MathML 5 through the process.

From Deyan to everyone: I can use the members mailing list when I write to
folks, so that everyone is aware they've been contacted ?

PL: have 30-minute meetings instead of an hour-long meeting.

NS: We have a new member.

NS: People should think up more topics to work on.
<https://cryptpad.fr/#cp-md-0-zoom-math-wg-intent-meeting-summary-2026-01-09->Zoom
Math WG Intent Meeting Summary (2026-01-09)
<https://cryptpad.fr/#cp-md-0-summary>Summary

The MathML Working Group held their first meeting of 2026 to discuss
ongoing specification updates and review recent progress. David shared a
pull request adding the 'A' element to MathML Core, which sparked a
detailed discussion about attribute inheritance and linking behavior
between MathML Core and Full specifications. The group also addressed
concerns about the W3C validator's handling of the 'intent' attribute and
discussed potential outreach to new browser vendors like Servo and Lady
Bird. Moritz provided an update on Wikimedia's plans to deploy MathML
rendering, though performance concerns with MathJax were identified as a
potential blocker. The conversation ended with a discussion about future
work priorities and the possibility of adjusting the meeting schedule to a
30-minute format, with meetings potentially being cancelled if there's no
new content to discuss.
<https://cryptpad.fr/#cp-md-0-mathml-an-element-discussion>MathML An
Element Discussion

The team discussed a pull request that adds an A element to MathML, with
David explaining that the A element would have identical behavior to HTML
and SVG A elements, including attributes like href, target, and download.
They debated whether mrow in full MathML should have the same capabilities
as A, with David suggesting that if they want to keep people using mrow in
full, it should have all the capabilities of an A element. The team
identified issues with the current documentation, including misspelled
words and incorrect statements about href support, and agreed that they
need to clarify how MathML works before making any changes.
<https://cryptpad.fr/#cp-md-0-mathml-specification-updates-and-future>MathML
Specification Updates and Future

The group discussed several MathML specification updates and issues. David
will create a pull request for the MathML core specification and make
comments in issue #555 regarding attribute inheritance. Neil needs to
clarify mstyle inheritance behavior. The team reviewed progress on MathML
implementation, with Moritz providing an update on Wikimedia's transition
to MathJax, though performance concerns remain. David will investigate
updating the W3C validator to allow the intent attribute. The group
discussed future work items, including adding support for round hand and
chancery script variations, and potential outreach to new browser projects
Lady Bird and Servo. They agreed to maintain weekly meetings but may reduce
duration to 30 minutes if no significant items are pending review.