Am Do., 30. Juli 2020 um 19:03 Uhr schrieb Wolfgang Corcoran-Mathe
<xxxxxx@sigwinch.xyz>:

> I think it's ready, unless Marc or John would like to add anything to
> the previous thread.

I haven't found the time to take a look at SRFI 196 in detail, so I
could report only on the things that attracted my immediate attention.

What I would like to see is an extension of SRFI 196 to more general
algorithmic sequences. Let me dub them "progressions". Ranges would be
a subtype of (finite) progressions. For finite progressions, we would
have procedures like (progression-reverse PROGRESSION) whose
implementation depends on the type of progression. For example, there
would be a conversion procedure generator->progression.
progression-reverse called on the result of such a conversion will
have to use an O(n) algorithm, while for ranges, it would be O(1) as
in SRFI 196.

This will also fill some things that cannot be realized with the SRFI
196 ranges, e.g. progression-map, for which range-map is missing.

Marc