For large graphs, vectors are much more efficient than lists (both in
locality and in size).  To increase the applicability of SRFI 234,
graphs should be encoded as vectors (or, possibly even better, by some
procedure).

This may not be relevant for a "toy" application of SRFI 234,, but
relevant when one wants to implement an application for huge graphs
using Scheme.

(R6RS's hashtable-entries returns vectors and not lists by the same reasoning.)

Am Di., 22. Nov. 2022 um 08:09 Uhr schrieb Lassi Kortela <xxxxxx@lassi.io>:
>
> This draft is over 100 days old.
>
> IMHO (topological-sort graph [=]) is OK as is.
>
> Is the spec written correctly?
>
> "Each connection is a list of the form (node dependency dependency ...),
> meaning that dependencies are dependent on node." -> Is this the
> customary order to write the graph? Makefile and similar syntaxes uses
> (node dependency dependency ...) where node is dependent on that
> dependencies. I'd anticipate this is more convenient for most
> applications. A utility to invert the graph would be needed.
>
> "It is an error if the same node (in the sense of =) appears in more
> than one connection." -> Appears at the head of more than one connection?
>
> Another procedure that returns each
> https://en.wikipedia.org/wiki/Strongly_connected_component as a list
> should probably be added.
>