On Fri, 6 Dec 2002, Stephen Pair wrote:

> But, if you require it as a subclass before you'll consider using in
> Seaside, then I'll make that change...better to have one solution to
> this problem than two; having two different solutions is definitely a
> maintenance issue.

Very true.  Thank you.

It's clear the next release of Seaside will need to be as a "load script"
- I'm currently in the process of breaking out Continuation and some
related goodies into a Backtracking package which Seaside will depend on;
Backtracking will in turn end up depending on IVAP.

The thing I'm playing with backtracking for now is an implementation of
John McCarthy's "amb" operator, which allows for some pretty neat control
flow games... more on that later.