RE: [EXTERNAL] Re: a minimal core intent proposal

Yes, our Unicode list consists of default names. The names can be overruled by using intent. Very interesting that other folks have used ℘ for things other than elliptic functions. It’s a good looking symbol 😊 But I really don’t think “script capital p” is a good name even in an intent string since ℘ isn’t a capital. And calling it script confuses it with the math script and math calligraphic p’s. Maybe var script p. Slightly analogous to varepsilon (ε) vs epsilon (ϵ).


From: Deyan Ginev <>
Sent: Thursday, November 10, 2022 3:56 PM
To: Murray Sargent <>
Cc: Neil Soiffer <>; David Farmer <>;
Subject: Re: [EXTERNAL] Re: a minimal core intent proposal

Hi Murray and Neil,

I agree that we should assemble a complete list with appropriate Unicode names, and try to keep it sterile from possibly domain-specific assumptions about what a character may or may not be useful for. Neil used the nice tagline of "self-describing", which would be a nice aim. I'm happy to offer help in doing the manual labor collaboratively here.

Nevertheless, I have to disagree about where we draw the lines of what should and should not be assumed by convention.

If a ℘ is used to mean "weierstrass-p", it really needs an intent annotation.
"script-capital-p" is the appropriate baseline for higher-education documents without any intent markup, such as the current state of the arXiv HTML5+MathML corpus.
Some examples:
- in arXiv:0812.1728, the variable ℘ is a "nonempty collection of subsets"
- in arXiv:1904.08618, the variable ℘ is "an irreducible polynomial of positive degree"
- in arXiv:2110.06614, the sequence of variables in ℘_1 to ℘_s are "prime ideals".

Stephen W. offered a lovely list of examples where an msup is not a power. I will also remind that there are about 40 examples of msup notations that different from "power" in the current Intent Open spreadsheet.

Any set of defaults that over-asserts its assumptions will be considered as non-standard in my eyes. Such notation grammars are certainly useful, but belong under a tight umbrella headings. For example, "℘" defaulting to "weierstrass-p" is reasonable only in an explicit context of "elliptic functions".


On Thu, Nov 10, 2022 at 6:36 PM Murray Sargent <<>> wrote:
Re speech for Unicode math symbols, their Unicode names provide a useful start and may be what we want. But due to stability considerations (Unicode names cannot be changed), some names are not what we want. A glaring example of this is the Weierstrass p (℘) which has the unfortunate Unicode name “SCRIPT CAPITAL P”. It’s neither upper case nor script; it’s its own thing, namely Weierstrass p or \wp in TeX. Amusing piece of history: in our laser theory papers and books, my physics colleagues and I used ℘ as the base of electric-dipole matrix elements and called it “squiggle”. It’s #296 in the American Institute of Physics Style Manual.

I can supply the English names of 360 math symbols that our software speaks. There are quite a lot more Unicode math symbols, but they are pretty obscure. It’s probably worth creating a complete list. I think I can get the names for about 18 other languages as well in case we want to get into localization. I also have names for the ~1000 math alphanumerics.


From: Neil Soiffer <<>>
Sent: Thursday, November 10, 2022 2:55 PM
To: David Farmer <<>>
Subject: [EXTERNAL] Re: a minimal core intent proposal

We discussed this some at the meeting this morning. My proposal included a statement that there should be a list of common intent names to help authoring software know what to do. If is kind of buried at the start of the document, so I've added a section calling that out more clearly. For Unicode, it does mention that we should provide a default speech (actually "meaning") table for Unicode chars typically used in STEM documents.

I have also added an "Internationalization" section that raises the question of  who/what should be responsible for internationalization of non-core intents. That includes non-core concepts (in my proposal) for things like "absolute-value". I suspect this might be a hot button topic.


On Thu, Nov 10, 2022 at 4:26 AM David Farmer <<>> wrote:

Let's consider absolute value.

Is it this document, or somewhere else, that tells me to use
the string "absolute-value" when I am specifying that?

|X| can also mean determinant, cardinality, order, or length.
Maybe not all of those are K-12 or K-14, but I think those should
all be in core because they are reasonably common.  I am suggesting
that as a general principle.

Similarly, the draft mentions (a,b) as an open interval, but it
could also be an ordered pair or a point in the Cartesian plane.

Not sure if you are asking for specific instances, but one of my
go-to examples is the LaTeX \times, which in K-14 can be
cross product
"by" as in 3-by-3 matrix or 10-by-12 foot room.

A bit past K-14 it can be direct product or cartesian product,
so by the principle I suggest above, those intents should also
be in core.

In biology, × is used to indicate a hybrid of two species, but
maybe we don't care about that.

On Wed, 9 Nov 2022, Neil Soiffer wrote:

> I wrote a proposal for simplifying what goes into intent core. It ended up being sort of an "AT requirements"
> document for core. If I extend it a little further to include what AT should do with "intent" (currently just
> presumed everyone knows), it would be the basis for an actual AT requirements document (or appendix). It also
> serves to let authors/authoring software know what they can count on as default behavior by AT.
> The proposal contains some open questions, but I believe it is fleshed out enough that it is understandable and
> actionable (let's do this/don't do this). It extends what I put in Deyan's intent spreadsheet and also has
> explanations. It will be the basis for the third agenda item on Thursday.
>     Neil

Received on Friday, 11 November 2022 00:35:58 UTC