[MLton] Successor ML and MLton

[Matthew Fluet]

> I've been playing a little with Andreas Rossberg's HaMLet-S
> (http://www.ps.uni-sb.de/hamlet/#successor-ml) and I like many of the
> simpler, mostly syntactic, extensions a lot (like the record construction,
> pattern matching, and module language shorthands).
> 
> So, I've been wondering how people feel about supporting those in MLton.
> Any thoughts?

I think it would be great to support the Successor ML extensions.  The 
MLB annotation system provides an easy way of indicating what source 
files should be elaborated with the extensions.  (And, using 
"-default-ann 'successorML false' -disable-ann successorML", one would 
get an error if any included source file used the extensions.)