[MLton] improved overloading for SML

[skaller]

On Sun, 2005-10-16 at 15:56 -0700, Stephen Weeks wrote:

> The only drawback of this approach w.r.t. the current ad-hoc approach
> is that there is no automatic constant overloading, and so constants
> require some additional syntax.  I personally don't find that very
> debilitating, since most constants are named and not sprinkled
> throughout code.

FYI Felix does overloading, and the one 'constant' which requires
annotation which is a pain there is the empty list (but then, there
is no HM inference at all, only synthesis, so MLton might not
pay this price .. OTOH Felix overloads named functions, not just
operators, the price being a requirement to annotate function
arguments)


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