]> | You basically want O(1) for set equality, I suppose. Actually, no --- perhaps I should have made clearer what I *really* want. The efficiency of comparison wasn't my motivation, but rather elegance and aesthetics. And I meant "canonical" with respect to ordinary structural equality, not necessarily pointer equality, so the problem is potentially a bit easier than you might have thought. I want to be able to treat an abstract type in a truly abstract way, and not worry about special-purpose equality relations on certain types. Otherwise it's an ugly mess dealing with complicated nestings like sets of pairs of lists of sets. Now, I think the right solution, conceptually speaking, is to allow user-defined equality on abstract types. But as far as I know this cannot be done in OCaml, and I've never met much enthusiasm for the idea among the CAML or SML experts. So a poor second best is to define abstract types in a canonical way, which was the starting-point of my question. After your remarks and Brian's, I'm starting to wonder if it is possible at all to do what I want. Maybe I should be looking for an impossibility proof instead... John. ------------------- To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/ Beginner's list: http://groups.yahoo.com/group/ocaml_beginners