🚀 go-pugleaf

Olivier Ridoux wrote:
> Moral: rational term unification and conservation of sharing are
> rather independent issues.

Sure, but I didn't claim that rational term unification and conservation of
sharing are dependent, rather that making the built-in predicates support
cyclic terms, leads to conservation of sharing - and of course this is not
strictly true either, because one has always the freedom to mess things up.

E.g. if one makes copy_term/2 such that it can make a finite copy of a cyclic
term, one can as well make a copy with the same internal sharing as the
original term - and in fact I think one has messed things up if not.

Bart Demoen