> > Could you provide a reference to de Bruijin indexing?

Xavier kindly gave a nice explanation.:) Thanks!

> On a related note, how would that compare (performancewise) to an
> approach like "abstract syntax" (represent a function not
> as (<id>, <exp>) and neither as <exp> (as in the case of deBruijn)
> but as fn x => <exp>) ?

As far as I know, to use that "higher-order abstract syntax" approach,
we have to _translate_ a GML program into a Caml code and evaluate it
at runtime.  That is exactly what I did in the "compiling" version
(PLClubCN) of our entry, but it didn't work well because the
compilation itself (including ocamlopt.opt) took some time, which
didn't pay.

Eijiro