On 2021-04-02 21:22 +0200, Marc Nieper-Wißkirchen wrote:
> Even if the extra allocation is insignificant, its existence shows that
> using Maybes is not the most direct approach but uses one unnecessary
> indirection.
>
> Another reason that may be compelling is the very existence of multiple
> values in Schemes. Maybes instead of continuations is like returning lists
> or vectors instead of multiple values.

OK, that makes sense to me.  Generating (unreified) multiple values
is indeed useful, and the Maybe protocol I suggested is really just
a minor variation on generating vectors.

Thanks for explaining.

--
Wolfgang Corcoran-Mathe  <xxxxxx@sigwinch.xyz>

"It is a syntax error to write FORTRAN while not wearing
a blue tie." --James Iry