On Sat, Oct 25, 2003 at 01:55:56PM -0700, Bradd W. Szonye wrote: > Bradd wrote: > >> That's a big pitfall! A right-fold on an infinite sequence will > >> diverge before the folding function ever gets called. I think it's a > >> bad idea to provide an interface that is "broken by design" like that > >> -- it cannot possibly function correctly. > > scgmille@xxxxxxxxxxxxxxxxxx wrote: > > Its not really. collection-fold-right will raise an error if applied > > to an infinite collection. Would it placate you if it was required to > > raise an error with truely infinite collections? > > That's better, but I still have a problem with the fact that you're > requiring something that is not well-defined. And by "not well-defined," > I don't mean that it's ambiguous or unclear -- I mean that there's no > way to define it meaningfully. But its far less elegant to make specific exceptions. After all, you're arguing over *what kind* of error occurs if someone uses collection-fold-right is applied to an infinite collection. Scott
Attachment:
pgpeHL5MwWGyo.pgp
Description: PGP signature