MathML-Core Meeting Minutes from July 26, 2021

7/26/2021 MathML Core Meeting

## Attendees:
* Neil Soiffer
* Louis Maher
* Paul DeMarco
* Deyan Ginev
* Brian Kardell
* Bert Bos
* Manuel Rego
* Stephen Watt


## Announcements/updates

#### FPWD update
https://mathml-refresh.github.io/mathml-core/
now redirects (thanks DC)

#### Some development progress updates
BB: has put out the request to publish the FPWD.

#### TPAC: any groups to meet with?
BK: TPAC virtual.  want to do liaisoning. people want joint meetings. We
need to have  specific plans to introduce ourselves to other TPAC
committees.

NS: introduce ourselves to ARIA if we want to get intent into MathML.  We
must get documents we can present to ARIA.

NS: Do we want to talk to CSS?

BK: Core may not have anything for TPAC.

BK: Neil has moved the issues from one repo to the other.  Labels were lost
during this process.

BK: We need to fix up the labels. Everyone  has permission to add labels.

NS: We should create the labels before the meeting so that people can think
about the issues.

BK: We will go through the issues to see who we should be meeting with. We
can correct labels during this review process.

## Agenda
NS: some commits are starting to land in the chrome branch again - rego and
fred landing things....Rego has added table support and Fred is working
stretchy chars.

BK: Suggests that we use our time today to triage issues that neil moved
over, and sort out labelling strategy on them, specifically with 3 things
in mind: 1) What is level 2; 2) what is a good agenda item for next month
3) thinking about what we want to organize for TPAC liaisoning .

WE want to put on a TPAC label if we want to discuss this issue with other
TPAC committees.

BK: Fred's  two new issues (https://github.com/w3c/mathml-core/issues/102
and https://github.com/w3c/mathml-core/issues/98) should stay open. Do  not
have a labels.

NS: https://github.com/w3c/mathml-core/issues/45 - If core spec have
pictures should the pictures look proper or does it matter? Should this
picture be labeled. The labels were lost during repo transfer.

BK: Found examples that should be updated earlier as well.  This issue
contains a sample rendering of MathML about CSS, the rendering of the
MathML is not quite correct. The example itself is "correct" in that the
example discusses font-family and color and these are reflected, but NS
wants better pictures. BK agrees better example renderings are helpful,
while not normative - and FW (on the issue) agree that these are
non-normative and therefore should be updatable whenever.

BK: https://github.com/w3c/mathml-core/issues/46 update core this might
bring new unicode scripts.
This is non normitave.

BK: https://github.com/w3c/mathml-core/issues/47
font enumeration in core.  Nothing in the spec really enables font
enumeration, this is CSS stuff.  I don't think we have to talk to the
privacy group at TPAC (there are some other issues too), but maybe fonts
group.

NS: wants to talk to the fonts group.

BK https://github.com/w3c/mathml-core/issues/48 is also from the same
person, I guess the ask is to split apart  privacy and security issues.  We
have a privacy and security section and it's also in the TAG issue and
explainer - I suppose this should like to issue 49 and 50 from the same
author as the ask is to separate these into 2 sections.

BK: https://github.com/w3c/mathml-core/issues/38 "add new method to the
MathML interface."" Not for level one. This is level two.  NS, can you
describe it?
NS: In writing polyfills get the hight and depth relative to the baseline.
This information is hard to get with out having hacks.  This information is
available if you can access it.

DG: wants this issue renamed.
BK: would like to also not prescribe a solution as much as describe the
problem - probably liaison with CSSWG on this

DG: has some access to change labels. added "label 2" to #38

DG: wants short labels.

Put TPAC labels on some issues that we want to talk to others about.

BK: https://github.com/w3c/mathml-core/issues/39
Script size multiplier: it is in MathML full but not in core. Useful for
people with low vision to control this.  Helps low vision people see
superscripts and similar small print items.
This gets a TPAC label.  did not go to CSS.
Keep this in level 2.

BK: Try to not add things to level one by default - especially in the first
meeting.

DG: What is in MathML that is not in core?
BK: Things in MathML full may not be in core.
NS: level 2 is as important as is level one, but level two will be
implemented later.

DG: level two is level one with higher difficulty?
BK: Not exactly. At this point anything with difficulty that isn't in level
1 already would be difficult to add - but the point of levels (as in CSS)
is to focus us on particular details and agreements to increase progress
and complete.  The CG did a lot to identify what was most important in a
somewhat achievable scope that we should focus on first.  There is a lot
already that we know needs to be tackled next. That would be for Level 2.

NS: issue 40 what character should be used for accents and hats.  Should
not be in core.

Talk to the fonts group in CSS.
Apple is interested in this.  There are numerous linked CSS issues here,
though they are not currently linked through this issue.  Add tpac label.

CSS handles fonts.

issue #42 move to level two.

renaming property currently level one.

#44 attributes not mentioned in core. Question for Rego.

RC: working on horizontal centering in a column.  keep this in level 2.

NS: #44 for tpac about decimal alignment. If you have numbers in a column
you want to align the decimal points.

BK: is there a taskforce of CSS to work on this alignment problem?

SW: If you have a column you can left or right or center align each column.
 you must do this for each item separately.
You want one way to align all the decimal points in a column.

#52 NS: says this is critical.  stays at level one.  Fred and the group
must reach an agreement.

#53 NS: thought that this was resolved and is now level one.

#63 tab navigation and links is in level 2.
#69 level two.

#71 handle unknown mathML elements. It is level two. not is core.

BK: asks people to look through the issues and bring problems up in emails.

NS: Lots of issues for ARIA - can we use ARIA label?

BK: Is there a reason you cannot?

NS: ARIA label difficult to use Braille. You cannot generate braille from
text. ARIA labels are meant for interactive tags.  Math is not an
interactive element.

BK: But those things are not unique to MathML, I guess I was suggesting AAM
is where to focus a TPAC session maybe, a subset of ARIA folks, because it
will force the intersection of those issues.  MathML AAM is new.


BK: use the agenda label if you want to discuss something next time.


NS: encourages people to add appropriate labels.
NS: the "needing test" label is important.

-- 
Brian Kardell :: @briankardell :: bkardell.com