bug#69709: `sort` interface improvement and universal ordering predicate

[Daniel Mendler]

Mattias Engdegård <mattias.engdegard@gmail.com> writes:

> 29 mars 2024 kl. 12.38 skrev Daniel Mendler <mail@daniel-mendler.de>:
>
>> I would
>> like to back-port the new feature in the Compat library, specifically in
>> the emacs-30 branch of the Compat repository. I've seen you mentioned
>> some pure Lisp code, implementing the new feature, in a slower, but
>> backward compatible way, which we could perhaps use?
>
> Here is my proof-of-concept hack that maybe you could use as starting point.
> It's incomplete but I think you can fill in the blanks. You could use the 
> tests
> in fns-tests.el, perhaps adapted to fit. Let us know how it goes.

Just letting you know that I've implemented value< and sort in the
emacs-30 branch of Compat. It works well so far, but value< is not yet
completely implemented. See the following commit:
https://github.com/emacs-compat/compat/commit/8190769d9eb9258dd8361bd322d90228dc586770

There is one thing I'd like to ask about value<. Would it make sense to
support comparing mixed types, e.g., numbers and markers or strings and
symbols? In particular I often like to mix numbers and markers, in
particular in cases where we can avoid creating a marker for performance
reasons if the pure position is sufficient and the location doesn't
change due to edits.

Daniel