[MLton] ML Workshop, sML Evolution, and CUFP trip report

[Stephen Weeks]

> Well, you missed the obvious:
> 
>    SML/NJ: Matthias Blume, Dave MacQueen, John Reppy

Whoops.  Nothing subconscious there.  That does remind me of another
point, which John Reppy brought up.  He felt that the lack of a common
build system was a major impediment.  That's somewhat understandable,
since he's been doing development with both MLton and SML/NJ lately.
I mentioned that I think there are bigger obstacles to portability
(language differences, library differences, FFI, ...).

There was some discussion about the possibility of a common underlying
system, e.g Matthias Blume's portable library graphs 

  http://mlton.org/References#Blume01_02

One thing I've always found problematic about this idea is that one
needs to translate both from the underlyign system to the high-level
notation (CM, MLB, SMLSC) and vice versa.  I don't see that working
very well with the fact that the high-level notation is intended to be
read and written by programmers.