Meta reply scgmille@xxxxxx 12 May 2003 20:29 UTC
Sorry for the delay in responding to so much commentary.  I think Oleg
has an excellent case for enumerators as the primary collection
traversal mechanism.  Oleg: I'd prefer a separate fold operator
(collection-foldl, for example).  This would give implementors the
opportunity for more under-the-hood work on efficient folding for
collections.

Also, I do feel that the current spec inhibits the creation of
functional or partially functional (linear updating) datastructure
operations.  I'd like to hear some proposals, but initially, my thought
is to specify both purely functional and linear updating versions of the
operations that were previously destructive.

Thoughts?

	Scott

--