Re: [Texmacs-dev] [TeXmacs] A question and proposals about references

[Frank]

I guess that labeled environments are something desirable. I don't see any 
intrinsic meaning of labels without a specific environment.

This seems to be how ConTeXt does it:

> \placesubformula
> \startformula \startalign
> \NC a_1 x + b_1 y \NC = c_1 \NR[eq:c][c]
> \NC a_2 x + b_2 y \NC = c_2 \NR[eq:d][d]
> \stopalign \stopformula
> (\in[eq:c]) and (\in[eq:d]) form a linear system of equations.

and

> \starttheorem[thm:pythagoras]{Pythagoras Theorem}
> The square on the hypotenuse is equal to
> the sum of the squares on the other two sides.
> \stoptheorem

Best,
Frank

On 4/10/23, Massimiliano Gubinelli wrote:
> I agree, but I think this is an early design problem given by following too 
> tightly the LaTeX model. It is not obvious how to change things and not break 
> old documents.
>
> Anyway I think a good solution to this problem would really improve a lot the 
> design of the document format.
>
> best
> Max
>
>
>
> > On 10 Apr 2023, at 17:32, Frank <sci.frank@gmail.com> wrote:
> >
> > Dear all,
> >
> > Let me first quote a paragraph in TeXmacs documents:
> >
> > > You first have to create a label at the position to which you want to refer. 
> > > For equations, the label should be created inside the equation, and similarly 
> > > for theorems and most other tags. The labels for sections, subsections, etc. 
> > > should rather be positioned just after the section titles; this prevents 
> > > TeXmacs from duplicating the labels in the table of contents.
> >
> > I did not find any rationale for this. Now I look at some codes, and if I 
> > understand correctly, it is doing the following:
> >
> > (set-binding name) — setting a binding with name "name" (this seems to be 
> > undocumented).
> > (label lab) — setting a label lab, which has the *last* binding.
> > (reference lab) — a reference to the label lab, displayed as the name of the 
> > associated binding.
> >
> > I find this counter-intuitive — where we set a label, it is a priori unclear 
> > where is the last binding, plus depending on a very low-level undocumented 
> > machinery called binding. For example, we can look at the following artificial 
> > TeXmacs Scheme code:
> >
> > (document (theorem (document "Consider the following conjecture" (indent (document (conjecture (document (concat "Some conjucture for " (math 
> > "n<in><bbb-N>") "."))))) (concat "It holds for " (math "n<leq>100") "." (label "thm:1")))) (reference 
> > "thm:1"))
> >
> > The label "thm:1" is associated to the binding produced by the "Conjecture 2", with name "2" instead of 
> > "1". This does not seem to be intuitive. This could be resolved by moving the label "thm:1" before "Conjecture 
> > 2", but this does not seem to be good for me.
> >
> > I would propose the following:
> >
> > 1. "Bindings" should be local to the environment. In particular, in the preceding code, "thm:1" 
> > should refer to 1 — the binding in "Conjecture 2" should be local in this environment, and outside this 
> > environment, the value should be resumed to be the one of "Theorem 1" environment.
> > 2. We should have abstractions "labeled-theorem", "labeled-conjecture", which 
> > allows others to avoid low-level abstractions such as bindings.
> >
> > Best wishes,
> > Frank
> >
> > _______________________________________________
> > Texmacs-dev mailing list
> > Texmacs-dev@gnu.org
> > https://lists.gnu.org/mailman/listinfo/texmacs-dev
>
>
> _______________________________________________
> Texmacs-dev mailing list
> Texmacs-dev@gnu.org
> https://lists.gnu.org/mailman/listinfo/texmacs-dev