>
> No. Compiler trates polymorphic types as abstract then. It need not 
> know
> what is the exact type, all it cares about is size of data, which in
> case of OCaml on 32 bit machines is always 4 bytes.
>
> For example in:
> .......

Fair enough. I'm curious though: Say I'm the programmer and I would 
like to solve the problem of combining different types of numbers with 
a fairly uniform syntax. I'm either going to define a union type for 
numbers and have all my functions pattern-match, or else I'm going to 
write several versions of my functions to handle different types of 
numbers and call different ones as appropriate. I would wish that if 
all my functions were built up from the primitives +, -, *, /, etc. the 
compiler could do this for me. The compiler may have to generate many 
different versions if I use them all, but otherwise I would have to 
write them myself!

I don't think this whole thing is a good idea in general, but I don't 
buy this practicality/efficiency argument against it.

>
> Again: Haskell does it, so typing isn't real issue here.

Haskell also does not have mutable data, and I think it has been well 
established in the literature that mutable data complicates subtyping 
and intersection typing. So, as others are pointing out, I definitely 
think typing is a real issue here.

-Mike

-------------------
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