[MLton] List comprehensions?

skaller skaller at users.sourceforge.net
Thu May 31 17:46:05 PDT 2007


On Thu, 2007-05-31 at 13:49 -0500, Matthew Fluet wrote:
> Andreas Rossberg wrote:
> > Wesley W. Terpstra wrote:

> On the other hand, higher-order modules (without first-class modules) 
> shouldn't be a problem.  (Again, the difficulty is teaching the 
> type-checker new tricks, but the target language would support it.)

That's not a real difficulty .. the REAL difficulty
is teaching programmers how to use them.

I have a ugrad degree in pure math with specialisation in
abstract algebra including category theory, and 30 years
programming experience .. I've even written my own compiler.
And I still really don't understand how to use functors in
Ocaml for anything non-basic.

This is basically a theoreticians problem: the current
representation of basic categorical concepts (modules,
functors, etc) and the interaction with ordinary data
type polymorphism is very clunky -- the underlying theory
is quite beautiful, but applying it in non-trivial
circumstances is not well understood.

-- 
John Skaller <skaller at users dot sf dot net>
Felix, successor to C++: http://felix.sf.net



More information about the MLton mailing list