no-monomorphism-restriction.md

@@ -73,6 +73,22 @@ There have been 4 other suggestions related to the monomorphism restriction:
 >
 > Seems against the spirit of Haskell.  But see [ Let should not be generalised](http://research.microsoft.com/~simonpj/papers/constraints/index.htm) which argues the case for this approach, at least for nested bindings.
 
+## Opposition
+
+
+In the discussion on the mailing list, Simon Peyton Jones wrote:
+
+
+As many of you will know I am now leaning 
+strongly towards doing no generalisation whatsoever on nested bindings.   See   
+[ http://research.microsoft.com/\~simonpj/papers/constraints/index.htm](http://research.microsoft.com/~simonpj/papers/constraints/index.htm)
+
+
+So I'd cut the cake differently:                                                
+
+- top level: generalise, no monomorphism restriction                    
+- nested: do not generalise, monomorphism restriction for every binding 
+
 ## References