This has nothing to do with the thread, but Keith's comments:
> then mytake is now implemented as a function of *three* arguments: the
> number and the list, as before, but also a `dictionary' which looks
and:
> [it's no accident that class constraints `C a =>' are written the way
> they are... they are really extra arguments `CDict a ->'.]
remind me of a Mark Jones idea from a few years ago (PEPM'94; Lisp &
Symbolic Computation 8, 3, 1995) to use partial evaluation to achieve
overloading without dictionary arguments.
Is this idea sitting on a back burner somewhere or was it abandoned due
to problems with separate compilation and code explosion?
Clem
