Minutes: MathML Core meeting: Sept 14, 2020

MathML Core Meeting of August 31, 2020

Attendees:

   -

   Neil Soiffer
   -

   Brian Kardell
   -

   Louis Maher


   -

   David Carlisle


   -

   Rob Buis
   -

   Bruce Miller
   -

   Murray Sargent

Regrets:

   -

   Fred Wang




Agenda:
https://github.com/mathml-refresh/mathml/issues/8#issuecomment-679384902

The meeting was recorded:
https://benetech.zoom.us/rec/share/l5IBEoUtpMwR2fKHeX-p20pNz1IeMHe3G-Bqm4dICSRJMqOyNjCJFLyggv9AE8s.qEiHaxj_oHFkbjG-
Passcode: r4%6He2a (starts about 7 min in).


Implementation updates

RB: Operator dictionary is finished and lspace/rspace is in chrome.

NS: That is a nice improvement in the appearance.

RB: Fixed some weird combinations of HTML and MathML and SVG. So
integrating foreign namespaces. Added more tests.

NS: I might be mistaken, but from the commit message it seems as if you
include SVG inside of MathML, it is absolutely positioned and so the math
won’t flow around it. Can you or FW check that. It seems like a problem.

RB: Next is operator layout -- stretchy or large op.

NS: Anything happening with table/row layout?

RB: Maybe in four weeks. Waiting until after Chrome implementation of
tables is finished.
CSS updates

BK: We have a meta issue with CSS WG:
https://github.com/w3c/csswg-drafts/issues/5384

BK: We went over this with WG and Igalia implementors. 5385 (math
display/inline) was resolved as proposed.

BK: text transform values 5386 wasn’t totally resolved. Math-auto (for
italics) was approved, but not for the math alphanumerics. Not sure how we
deal with compatibility between the various values.

BK:  Agreed to math-style (normal/compact). Didn’t get to math-shift and
scriptlievel.

BK: Will have another meeting. Scriptlevel will probably be the majority of
the time
Resolving more issues:Decide how to handle unknown MathML elements #139
<https://github.com/mathml-refresh/mathml/issues/139>

BK: I suggested perhaps we should keep in L1 for now in the past and review
if we need to move to level 2 later. It might be time to consider moving it
to L2.

BK: This is about things in the MathML namespace. There is about how the
parser handles them (unchanged), how they are rendered (as a generic
element), and there is the interface on how you can manipulate them. So the
proposal is to keep the interface as MathMLElement.

BK: SVG also has an unknown element and didn’t expose the proper IDL.
Didn’t get much buy in and removed it. Trying to align with SVG would be
good, but takes time so move to level 2.

DC: I’m happy to leave to Igalia whether they have time to do it for level
1 or not.

NS: I don’t think there would be much difference in the interface if we had
it.

BK: Right, not much difference. In L1 it would literally just be the
constructor name.  There might not even need to be a distinction as in HTML.

Consensus: punt to level 2 and try and coordinate SVG
Applying visual effects #179
<https://github.com/mathml-refresh/mathml/issues/179> -- Rob added some
tests -- status?

RB: didn’t want to add too many tests. There are 11 CSS filter tests. I
added four as a start. Seems to be handled well.

RB: Can remove all labels except ‘needs tests’.

Action Item: RB: I’ll ping google to see if there is some scenario that
needs to be tested.
automatic size adjustment: #225
<https://github.com/mathml-refresh/mathml/issues/225>

NS: Looks like we need to check with CSS WG as to what they think is best.

Action Item: BK to talk with CSS folks.

U+002D HYPHEN-MINUS in mo operators: #146
<https://github.com/mathml-refresh/mathml/issues/146> => write a polyfill?

MS: Writers should always use 2212, but we should be flexible for legacy
purposes.

NS: Probably ⅔ , maybe 80% of MathML existing uses it.

DC: Even the MathML spec uses ‘-’.

RB: The Igalia branch supports ‘-’.

MS: The unicode technical committee screwed up badly with the angle
brackets. They unified them with the Asian chars and then we added new
ones. 2329 and 232a are the old ones and they should be converted to the
new ones. MathType generated the old ones.

DC: I got the HTML entities to point to the right characters. The MathML
spec has it right. It could be done with a polyfill.

Action Item: BK to reach out to the google folks to see what they think in
terms of how one would implement this.