Re: Hashing a procedure object reliably

On Fri, Apr 11, 2025 at 11:34:37AM -0400, Olivier Dion wrote:

[...]

> I think the definition of equivalent here is that the pure behavior of
> the procedures are the same.  Say we're hashing `(lambda (x y) (+ x y))'
> and `(lambda (u v) (+ u v))'.  These procedures does not have equivalent
> sources, but I do have equivalent semantics [...]

This problem in its general form is, alas, undecidable. You might try to
restrict it in some useful ways, if I remember correctly.

Cheers
-- 
t

-----BEGIN PGP SIGNATURE-----

iF0EABECAB0WIQRp53liolZD6iXhAoIFyCz1etHaRgUCZ/lKsgAKCRAFyCz1etHa
RoRMAJ4vpc28jqKF+X/cFT8ZIAyqQnrEKQCcCj5XRJQ82bccAUBv4GK/CZ94rZo=
=gVP5
-----END PGP SIGNATURE-----