Minutes: MathML Core meeting, 14 Dec, 2020

Attendees:

   -

   Neil Soiffer
   -

   Brian Kardell
   -

   Louis Maher
   -

   Murray Sargent
   -

   Patrick Ion
   -

   David Carlisle
   -

   Bruce Miller


Regrets:  Fred Wang

Thanks again to Louis Maher for taking notes.

The meeting was recorded:
https://benetech.zoom.us/rec/share/rOFaABY2NDKWMgomI3UvtqoVZtXnWoYBEkcHZD2bPBIvtt0otdAh2TQ9OJprhd5F.7ttatUh7M6eEFFZt
Passcode: Z7.rq*q6 (Starts about 6:30 in)



1. Status updatesa) Charter

NS: Working its way through the W3C process. Currently in horizontal
review. APA has approved. Still waiting on others. Then it gets put up for
a vote among members.

NS: Charter will not be approved until February 2021.
b) Chrome Implementation

From FW:

   -

   Core spec is pretty stable.
   -

   Remaining github issues might need clarification and/or testing, but
   they are not blocking work.
   -

   No recent patches landed in Chromium.
   -

   Stretchy chars are a concern for Google - they want to do a rewrite of
   their LayoutNG minmax algorithm but that’s not going to start ‘before the
   end of 2021’.
   -

   BK: Had to shift resources away from MathML.  They have some new hires
   which might help. Need alignment with other browsers.  Would like to get
   help to pay for the work. Igalia paid for the work in 2020. They would like
   financial help. Tables are taking a long time.  Tables are almost done.
   Stretchy may be getting work arounds.
   -

   NS: Missing in chrome: Stretchy characters, tables, and script
   positioning.
   -

   DC: Can we ask for other people to help out financially?
   -

   Google is rewriting TableNG and they plan to ship Q1 2021
   -

   Igalia needs to find sponsors to pay for more work
   -

   Might be able to work on Firefox/Webkit interop or Chromium MathML a11y


c) Polyfills

NS: Almost all the polyfills are done. I may rewrite the elementary math
one because it currently targets HTML tables which makes placing it inside
of MathML pretty ugly.

NS: Remaining polyfils are mstyle, semantics, named spaces (thinspace, …),
and script shifts. Also tables need to be done but the core spec needs work
so not sure what the target will look like.

NS: Probably the biggest thing is tests -- we need to figure out a testing
platform/framework.

NS: Other issues revolve around incompatibilities of the implementations --
hopefully that shrinks over time.
d) Anything else?2. Work to be donea) What issues need to be resolved?

NS: there are 112 open issues

   -

   5 need resolution
   -

   67 have the label “Core”
   -

   42 have the label ‘Need spec update’ -- I suspect a review would find
   most are dealt with
   -

   51 have the label ‘Need tests’ -- I suspect many already have tests
   -

   18 have the label ‘Compatibility’ -- these need review as I’m dubious
   about this based on looking at a few of them.
   -

   15 have the label ‘need polyfill’ -- I need to do an update; probably 10
   are real


b) What needs to be finished in the spec?

From FW: not much

NS: not sure tables are correctly spec’d

NS: Major work is needed on the full spec

NS: Doing an example element case for presentation, e.g., <mfrac>, would
serve as template for later working through the other elements of MathML.
For instance, attributes could all be listed with pointers to Core for
details on most of them (perhaps essentially transcludable)
c) Getting work done on implementations (Chrome, Firefox, Safari) so what
is spec'd actually is implemented.

NS: Will Igalia work on Firefox and safari?

BK: Yes. Complicated because other companies are having layoffs.

BK: Is everyone on the call joining the working group? Yes.
d) Timelines?

NS: will work on more polyfills.

NS: Polyfills must be checked across more than one browser.

BK: Discussed fund raising.

NS: Wants us to start writing the full spec. Maybe a chapter every few
months.

NS: Meet in the second week in January 2021.