Spell-out the names of APIs? Artyom Bologov (01 Sep 2026 12:11 UTC)
Re: Spell-out the names of APIs? Hernán Ibarra Mejia (02 Sep 2026 09:23 UTC)
Re: Spell-out the names of APIs? Marc Nieper-Wißkirchen (02 Sep 2026 10:27 UTC)
Re: Spell-out the names of APIs? Artyom Bologov (02 Sep 2026 10:33 UTC)
Re: Spell-out the names of APIs? Hernán Ibarra Mejia (02 Sep 2026 15:43 UTC)
Re: Spell-out the names of APIs? Marc Nieper-Wißkirchen (02 Sep 2026 15:50 UTC)
Re: Spell-out the names of APIs? Peter McGoron (02 Sep 2026 15:58 UTC)
Re: Spell-out the names of APIs? Hernán Ibarra Mejia (02 Sep 2026 16:44 UTC)

Re: Spell-out the names of APIs? Hernán Ibarra Mejia 02 Sep 2026 15:43 UTC

On Wed Sep 2, 2026 at 1:33 PM EEST, Artyom Bologov wrote:
> Maybe make them operator-like instead? Something like <- for define/m?
> This will ensure they are catching the eye and are treated with
> deliberation. And it consistent with Haskell, I believe?

A lot of the literature/implementations use `<-` because of Haskell (see the
Prior Art section). I heavily dislike it because:

- If it is in procedure position, as in `(<- name mndi)`, then the arrow points
  the wrong way.
- If it is in infix position, as in `(name <- mndi)`, then it feels anti-Scheme
  (and yes, some people use this form).

In general `<-` is meant to be infix, and Scheme does not deal with infix
operators [1].

For all its verbosity, `monadic-define` gets the meaning across. Though I
initially dismissed it, `define-monadic` also has some merits, namely it starts
with "define".

I will think more about this. I am open to other suggestions.

Hernán

[1]: But, then again, there are mathy operators like `<` which are really meant
     to be infix.