Minutes: MathML Full Meeting, 13 Nov, 2025

 Attendees:

   - Neil Soiffer
   - Louis Maher
   - David Carlisle
   - Paul Libbrecht
   - Patrick Ion

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

   - Deyan Ginev

<https://cryptpad.fr/#cp-md-0-action-items>Action Items
<https://cryptpad.fr/#cp-md-0-agenda>Agenda
<https://cryptpad.fr/#cp-md-0-2-a-href-https-github-com-w3c-mathml-issues-1-issuecomment-3518606914-simplification-of-the-mstyle-element-issue-1-a->2.
Simplification of the mstyle element · Issue #1
<https://github.com/w3c/mathml/issues/1#issuecomment-3518606914>

*ACTION:* DC will make an issue allowing columnalign on mtable.
<https://cryptpad.fr/#cp-md-0-3-a-href-https-github-com-w3c-mathml-issues-247-247-spec-should-specify-what-char-to-use-for-accents-lines-a->3.
#247: Spec should specify what char to use for accents/lines
<https://github.com/w3c/mathml/issues/247>

*ACTION:* We should discuss issue #247: Spec should specify what char to
use for accents/lines <https://github.com/w3c/mathml/issues/247> in the
core meeting.
<https://cryptpad.fr/#cp-md-1-agenda>Agenda
<https://cryptpad.fr/#cp-md-0-1-announcements-updates-progress-reports>1.
Announcements/Updates/Progress reports

PL wants to bring the Table of Suggested speech Templates for Unicode
Characters <https://w3c.github.io/mathml-docs/unicode-speech/> into the
spec (issue 546 <https://github.com/w3c/mathml/issues/546>)

PL: Add text to the full spec that links to the Table of Suggested speech
Templates for Unicode Characters. Put the link in the full spec next to the
concept list.

NS completed this.

PL: The sanitizer works in Firefox with MathML, but while the sanitizer
works in Chrome, it does not work with MathML in Chrome.

DC: The pdf MathML reader does not work in Firefox, but he is hopeful that
it will work by Christmas.
<https://cryptpad.fr/#cp-md-1-2-a-href-https-github-com-w3c-mathml-issues-1-issuecomment-3518606914-simplification-of-the-mstyle-element-issue-1-a->2.
Simplification of the mstyle element · Issue #1
<https://github.com/w3c/mathml/issues/1#issuecomment-3518606914>

There was confusion over the name decimalpoint. It was used to align
columns, and it was used to specify the character that acts as the decimal
point.

This is for the full spec.

Rename decimalpoint.

We will call the character used as a decimal point "decimalpointcharacter".

From Patrick D F Ion to everyone: Decimalseparator

PI suggested decimalalignchar for the element used to align columns.

*ACTION:* DC will make an issue allowing columnalign on mtable.
<https://cryptpad.fr/#cp-md-1-3-a-href-https-github-com-w3c-mathml-issues-247-247-spec-should-specify-what-char-to-use-for-accents-lines-a->3.
#247: Spec should specify what char to use for accents/lines
<https://github.com/w3c/mathml/issues/247>

NS: What character do you use for a hat? There is no agreement.

NS: No character worked in all the browsers. Sometimes they would not
stretch.

NS: We have no way to tell people what character to use.

NS: All the browsers should do the same thing for the same font.

There is definitely a browser problem in that the same font displays
differently in different browsers. There may be font issues also. E.g.,
some fonts may have support to stretch one character but not another.

DC: Core has a non-normative section
https://w3c.github.io/mathml-core/#combining-character-equivalences that
sort of says these characters should be considered the same. It is not
referenced in the rest of the spec and lacks more text explaining its
purpose. Actually there are two tables; I can't remember why.

*ACTION:* We should discuss issue #247: Spec should specify what char to
use for accents/lines <https://github.com/w3c/mathml/issues/247> in the
core meeting.

DC: TeX has a table of examples of what each font does. See
https://tex.org.uk/macros/unicodetex/latex/unicode-math/unimath-symbols.pdf
<https://cryptpad.fr/#cp-md-0-zoom-meeting-summary-11-13-2025>Zoom Meeting
Summary 11/13/2025 (lightly edited) <https://cryptpad.fr/#cp-md-0-summary>
Summary

The team discussed various technical issues including link formatting,
Unicode space characters in mtext, and browser compatibility problems with
MathML and fonts. They addressed attribute naming conventions and
implementation challenges related to decimal alignment for mtables, while
exploring differences between mstyle and mrow. The group also discussed
font support and character mapping inconsistencies across different
browsers, agreeing to seek clarification from implementors regarding
specification updates and character implementations.
<https://cryptpad.fr/#cp-md-0-unicode-space-character-fix-discussion>Unicode
Space Character Fix Discussion

The team discussed a Unicode space character issue in mtext, where Neil
realized he had not properly fixed the white space handling in a previous
commit. The group agreed to update the specification to clearly state that
an mtext element with all Unicode space characters should be treated as
white space. Neil also mentioned he had completed a table-related task that
David thought was not implementable, and they decided to discuss this
further in the next core meeting with Eri.
<https://cryptpad.fr/#cp-md-0-pdfjs-issues-and-updates-discussion>PDFJS
Issues and Updates Discussion

The team discussed several updates and issues. David merged a PR with Neil
to link to Unicode speech and considered attending a PDF tutorial next
week. Paul reported that the sanitizer is not working in Chromium, only in
Firefox. Neil mentioned a potential fix for tagged MathML in PDFJS for
Firefox, but no updates were provided on this issue. David shared feedback
from James Taperson about PDFJS navigation bugs, noting that progress is
being made but it may take some time to implement. The team also briefly
discussed the Mstyle element simplification issue, which remains open.
<https://cryptpad.fr/#cp-md-0-decimal-point-attribute-rename-discussion>Decimal
Point Attribute Rename Discussion

The team discussed renaming the "decimal point" attribute to
"decimalpointcharacter" to avoid confusion between the attribute name and
the alignment value name. They agreed to keep the alignment name as
"decimalpoint" while changing the attribute name to "decimalpointcharacter"
to better reflect its purpose. The discussion also touched on the need to
update the MathML specification to include HTML attributes and potentially
add decimalpoint as a new attribute in a future version.
<https://cryptpad.fr/#cp-md-0-decimal-alignment-for-mtables-discussion>Decimal
Alignment for mtables Discussion

The team discussed the implementation of decimal alignment for mtables,
with David proposing to create an issue for allowing decimal alignment on
mtables.
<https://cryptpad.fr/#cp-md-0-mstyle-attribute-styling-discussion>mstyle
Attribute Styling Discussion

They explored the differences between mstyle and mrow, noting that mstyle
allows for setting default values for attributes like linethickness on
mfrac, which is not possible in mrow. Neil suggested leaving mstyle as is
and moving presentation-only attributes over, rather than making it fully
compatible with core.

The team discussed attribute styling in Mstyle and Core, focusing on line
thickness and compatibility issues. Neil proposed moving certain sections
in Mstyle, while David agreed to create a new issue regarding alignment.
The group acknowledged potential incompatibilities between Mstyle and Core,
with Neil suggesting the use of polyfills to address these issues. They
also touched on the challenge of determining which character to use for
accents and lines, noting the complexity of the situation.
<https://cryptpad.fr/#cp-md-0-mathml-browser-font-compatibility-issues>MathML
Browser Font Compatibility Issues

David and Neil discussed issues with MathML and font compatibility across
different browsers and fonts. Neil mentioned problems with character
stretching and font compatibility. They explored various font options and
characters, noting inconsistencies between Firefox, Chrome, and Safari.
Neil concluded that it is difficult to provide a universal solution for
character usage due to the varying behavior across different browsers and
fonts.
<https://cryptpad.fr/#cp-md-0-browser-font-stretching-compatibility-discussion>Browser
Font Stretching Compatibility Discussion

The team discussed browser compatibility issues with font stretching,
particularly focusing on how different browsers handle characters like
tildes and hats. David explained that Chrome only stretches characters if
the browser specifically requests stretched variants from the font, while
Firefox and Safari handle stretching differently. Neil suggested that both
browser and font implementations need improvement, and proposed involving
font manufacturers to ensure better support for stretching characters. The
team agreed that the specification should settle on a consistent approach,
with authors using standard ASCII characters for accents while browsers
handle the stretching logic.
<https://cryptpad.fr/#cp-md-0-math-character-mapping-specification-issues>Math
Character Mapping Specification Issues

The team discussed issues with character mapping in mathematical
expressions, particularly regarding the use of combining characters and
their implementation across different browsers and fonts. Neil and David
identified inconsistencies in the specification, specifically in Appendix
B, and agreed that more documentation is needed to clarify character
mappings. The group decided to bring this matter up in the core meeting to
seek clarification from Eri regarding the Chrome implementation and to
explore potential solutions, including the possibility of selecting a
single character for consistent use. They also acknowledged the challenges
of font support and the need to consider the limited updates to existing
math fonts.
<https://cryptpad.fr/#cp-md-0-mathml-font-support-discussion>MathML Font
Support Discussion

The team discussed font support for MathML characters, with David sharing a
table showing how various characters work across different fonts. They
agreed to add a link to this table in GitHub issue #247
<https://github.com/w3c/mathml/issues/247>, which specifies characters for
accents and signs. David mentioned ongoing work with the LaTeX Core project
to improve accessibility support in package authors' classes, particularly
for mchem expressions.