Janice Caron wrote: > And if not, can it be made to work? (That is, can we implement dup for > all primitive types?) We could, but my thinking hasn't progressed that far.