[racket] Puzzled about type inference

On Tue, Aug 05, 2014 at 08:53:42PM -0400, Matthias Felleisen wrote:
> 
> You guys came up with some wonderful ideas. 
> 
> I think this particular one is easy to implement when 
> the program type checks. But when it doesn't, what do 
> you show? 

When the program fails to type-check, does the type-checker just throw 
up its hands in despair because at the point where the contradiction 
becommes apparrent it no longer has any knowledge as to where in the 
source code it was working?  Is the assignment of types to expressions 
not known at all until it is complete?  Hindley-Milner analysis is of 
this kind.  Or does it assign types in a mostly bottom-up fashion with 
possibly some top-down influence, until it finds one expression that 
has a type incompatible with context?  In that case, there's still 
clear information to present to the programmer.

-- hendrik