MathML-Core Meeting Monday Jan 27, 2025 (Special make-up 1.5 hour edition)

[bkardell]

• Clarify the language around deprecated features of MathML Clarify the language around deprecated features of MathML #279

• How is inline stretch size constraint determined? How is inline stretch size constraint determined? #270 (and attached PR... Is it ready? Open comments..)

• Too much space around Too much space around /  #260 (was it closed by regenerate opdict with lspace=rspace=0 for U+002f issue #260 #265?)

• Consider support for the box-sizing property Consider support for the box-sizing property #257 - (do we need spec updates?)

• Superscript consistent baseline in rows. Superscript consistent baseline in rows. #255 - it has no replies since, it was opened in August

• Accent rendering with <mover accent="true"> Accent rendering with <mover accent="true"> #254

• Be more explicit about how to get a largeop glyph Be more explicit about how to get a largeop glyph #253 (no replies opened in August)

• Vertical alignment of large operators Vertical alignment of large operators #250 (and associated pull request from Harry Clarify layout of large operators algorithm #276)

• [interop] Text fragments + MathML [interop] Text fragments + MathML #240 (this was on the agenda in May but we have no notes)

[polx]

• Let us discuss MathML support in the HTML Sanitizer API #227 till the end too.

[NSoiffer]

Minutes:

Attendees

Brian Kardell
Bruce Miller
Bert Bos
David Carlisle
Deyan Ginev
Neil Soiffer
Paul Libbrecht

Regrets

Louis Maher

Agenda Items
Announcements/Updates

BK: Process for going to CR?

BB: I will look it up and prepare the transition.
Clarify the language around deprecated features of MathML

#279

NS: How does this interact with MathML full? What's the decision criteria?

DG: How about something at level 2 is not deprecated? It could be listed as experimental.

DC: There are things used outside of the web. If (e.g., mathbackground) is not put into MathML, then there is nothing to hang the hat on. CSS could make use of those attrs.

BK: These can be presentational hints. There can be stylesheets that can be provided. Not sure if they are part of the UA stylesheet.

DC: If those or polyfills aren't provided, then MathML 4 is a waste.

DG: If deprecated from both, then it is deprecated. If in full, but not in core, then it is confusing.

BK: We should add a conforming section and mark things that are presentational hints as non-conforming, and hence deprecated even though they are widely implemented. This is similar to HTML's "font".

Action Item: MathML full to take a look at what should be depecrated or listed as experimental.
How is inline stretch size constraint determined?

#270 (and attached PR... Is it ready? Open comments..)

BM: the spec doesn't really say where stretchiness applies. It should be clearer. It would be expected that the mpadded example should work.

NS: What about other situations like mfrac.

DC: Also table element widths/heights.

BK: We should have some codepen examples with the expected stretchiness.

DC: Make this level 2.

BK: TBD on level 2

Action Item: BM to write up codepen examples.
Too much space around

#260 (was it closed by #265?

Action Item: Closed.
Consider support for the box-sizing property

#257

(do we need spec updates?)

DC: I don't feel competent to edit the document because there is no clear place in the spec and don't know what to say.

Action Item: BK to try and figure out a resolution.
Superscript consistent baseline in rows.

#255

it has no replies since, it was opened in August

Action Item: NS to suggest using mmultiscripts. Full has additional attrs to help.
Accent rendering with

#254

NS: There are two separate things. One is using the accent attr. We had a discussion over this and Fred was very insistant that it should not pick it up from the mo, and hence the mover attr must be used. The other issue is what character to use.

DC: OpenType dictionaries differ on which characters stretch.

NS: It seems wrongs that users need to know what character is stretchy in what fonts. The font to use is resolved by the browser, so we should say char XXX is the character to use for a stretchy char and the browser should figure out what that maps to.

DG: Maybe make the character mapping as level?

DC: It would be really good to be able to use the ASCII char

Action Item: Make which character to use be level 2.
Be more explicit about how to get a largeop glyph

#253 (no replies opened in August)
Vertical alignment of large operators

#250 (and associated pull request from Harry #276
[interop] Text fragments + MathML

#240 (this was on the agenda in May but we have no notes)
MathML support in the HTML Sanitizer API

#227

BK: the sanitizer by default is "strict". In this mode anything that is a display hack that hides things is removed. There is a config attr that allows someone to override this by passing an empty config.

BK: Who is going to use the sanitizer and math? Those that do, they can add this simple workaround.

BK: There is still a lot more work on the sanitizer. The key is to make sure MathML is not different/special.

Action Item: Close because we shouldn't try to say MathML is "special" and is unsafe. There is a workaround via the empty configuration.