Proposed charter for rechartering

[NSoiffer]

This issue is for discussion on the proposed re-chartering of the Math WG.

The document is located in the mathml-docs repo .

[polx]

I quickly glanced through the document and it seems good to me.
I don't understand why polyfills are mentioned under "Other deliverables" but not in the scope. Maybe this is intentional?
I would say "LaTeX/TeX to MathML" in words.
thanks
Paul

[polx]

Do we have to be more explanatory about our need for more time?

[physikerwelt]

• I wonder if we want to discuss if Content MathML stays out of scope.
• I would also suggest to discuss a canonical translation of TeX to MathML (see also @polx)

[NSoiffer]

@polx: wrote:

I don't understand why polyfills are mentioned under "Other deliverables" but not in the scope. Maybe this is intentional?

The original charter didn't have that. I'm not sure it is needed, but I added it.

@physikerwelt wrote:

I wonder if we want to discuss if Content MathML stays out of scope.

Yes. We discussed that at an earlier meeting and it was clear no one has the energy to want to tackle a new spec at this time. There is however a "gather requirements" mention and that a WD might be created that incorporates changes based on deferred issues (which might content MathML issues).

@physikerwelt wrote:

I would also suggest to discuss a canonical translation of TeX to MathML (see also @polx)

I added a deliverable:

Suggested list of LaTeX/TeX macros (names/args) and their mapping to intent core concepts and properties

Is this what you are suggesting? There is already a bullet item about a LaTeX->MathML translator that incorporates intent.

[physikerwelt]

@NSoiffer I am still unhappy about excluding cmml. My feeling is that now that we have intent, there is no use for non-strict cmml. For example a non-normative the LaTeX -> MathML translation list that also includes intents but does not include cmml seems strange, if we continue to support it. Maintaining the OpenMath CD https://github.com/OpenMath/CDs/tree/master and the intents concepts seems already some work. Also for example the current working draft does already contain changes to CMML like removing previously deprecated elements as described https://w3c.github.io/mathml/#chg_contm.

At least for the mathml core list there are concept that correspond to each other such as

definite-integral

and

int,interval

both ways of expressing the same mathematical concept does not work well together, and the cmml construct

<int/>
  <interval><ci>a</ci><ci>b</ci></interval>
  ...

does not seem very intuitive. And in https://w3c.github.io/mathml/#contm_p2s_int we spend a whole section on how to rewrite that to strict cmml. (Here we use defint instead of definite-integral). From my perspective, we would benefit from cleaning this up. And also I think it would foster the uptake of the spec.

[polx]

@physikerwelt wrote: I would also suggest to discuss a canonical translation of TeX to MathML (see also @polx):

I really just meant to spell out things but if we have the energey, we could offer this. Can we set this as an optional goal?

maintain cmml

We could include a sentence such as "do regular maintenance" which would include removing deprecation and possibly mark some cmml elements as deprecated or reformulate. At this stage, I think, however, that we should not promise anything with cmml.

[physikerwelt]

I really just meant to spell out things but if we have the energey, we could offer this. Can we set this as an optional goal?

I would make it optional how long that list is. Things I would like to discuss is

1. how to use mo vs mi for optimal spacing
2. when to use mtable (equation groups vs matrices)
3. how to get displaystyle/textstyle right (e.g. mover vs msup, )

For intent, I think we want to have latex examples. The comments in table already provide a good starting point. I could do the fraction that can be expressed in texvc (Wikipedia).

[davidcarlisle]

We could include a sentence such as "do regular maintenance" which would include removing deprecation and possibly mark some cmml elements as deprecated or reformulat

I think this is covered by the charter text saying that requirements for mathml5 are in scope.

We are not planning additional deprecations or technical changes for mathml4
Recommend that issues are opened for any such proposals