3/6/19 Meeting MathML core minutes

Thanks to Murray for bowing to the pressure to take notes this time :-)

The recording can be found at:
https://benetech.zoom.us/recording/share/4HwhS552wDT9qnBWNDDzaUnlHCBUYfnhcDreSCFQY76wIumekTziMw

Lots of action items here, including a few for some who couldn't make the
meeting today. As a summary, there are action items for *Neil, Dani,
Murray, Fred, Rob*, along with some general calls for my stats. Please see
below for details.

*Topic: linethickness (#4) and mathsize #7*
Neil reported the MathPlayer values used for these.
Would like to get more input.
Murray said Office mainly ignores them except linethickness=0.



*Action: Murray to double check what Office doesAction: Ask Dani to check
what MathType Windows and/or MathType Web usesAction: Fred to check what
values Firefox uses*
*Topic: Keep in Core as mrow-like? #26 (comment) #70 (comment), #65
(comment)*
Reaffirm previous feelings
   Keep maction in core but render as first child only.
   Keep merror
   Keep mphantom
*Action: remove "needs resolution tag"*

*Topic: fallback values for math constants: #69*
Microsoft Office requires OpenTpe math font for math characters unless
formatted as embedded normal text. MathType allows other fonts to be used
for math characters.


*Action: Neil to find out MathPlayer default valuesAction: Murray ask
coleagues how OpenType MATH table constants were derived. cc Neil,
FrédéricAction: Ask Dani to check what MathType Windows and/or MathType Web
uses *
Note: add what is found to
https://mathml-refresh.github.io/math-constants.html (edit
https://github.com/mathml-refresh/mathml-refresh.github.io/edit/master/math-constants.html
)



*Topic: mpadded: prohibit vertical pseudo-units in horizontal attributes (
#81 )*Fred: mpadded difficulty when specifying width as a function of the
height.

*Action: get more statistics for mpadded attributes.*Bruce: LaTeXML
probably doesn't use them
Neil: Can TeX set a width as a function of the height?
Neil: Is there a polyfill way of implementing such mpadded elements. If
they're not used much, maybe drop the relative units. If used, we can't
take it out of full, but what to do for a polyfill?
Murray: Office get info from open type tables
Neil: mpadded's height refers to the element height not to a <math> height
default.
Murray: OfficeMath doesn't use it.
Rob: I was hoping CSS calc could do it, but it doesn't seem to be able to
do it.
Neil: MathJax 2supports mpadded and the pseudo-units



*Action: Neil to look at MathML mail list to see if he can find motivation
for relative width capability.Action: Fred to look at Texzilla and
eTextToAction: Frédéric to check if implemented by MozillaAction: Rob to
ask Brian if there's a way to specify relative widths with CSS (in issue
#81)*
Fred: What about named spaces?
Neil: we agreed to get rid of them in core



*Topic: Add rules to map from non-combining to combining accents #11*Mapping
from non-combining to combining accents may have already been implemented.
(Is implemented in OfficeMath).
Table to next meeting when hopefully David C is present to answer whether
it has been implemented.



*Topic: Remove mlabeledtr? #72*mlabeledtr may be needed for more
complicated kinds of equation numbering. Can CSS do it? Bruce: mlabeledtr
hasn't been implemented much. Bruce uses HTML tables for this. Layouts
aren't just math; they're mixtures of text and math, so need HTML. It's
more about accessibility and computablility. Want to display fragmented
combinations with semantics for accessibility. The semantic info wouldn't
necessarily be displayed.
Neil: MathJax supports it, so it might be used.
*Action: Neil to see if MathJax stats for mlabeledtr?*
Neil: Hard for Frédéric to implement mlabeledtr? Maybe use an ID and refer
to ID? Good question for Brian.


*Topic: Reset CSS rules on the math too (UA stylesheet): #34 #35 #36*Resetting
CSS rules on math. Want to ignore many of them inside <math>...</math>
Fred: we need to decide which if any should be reset or allowed on MathML
elements.
Neil: someone may want to use absolute positioning instead of mpadded to
adjust things
Fred: or use borders instead of menclose
Neil: time is up and people haven't thought about this much recently. Let's
make it a priority for people to think about this for the next meeting.