Re: [Haskell] Type inference for arbitrary rank types

2006-06-23 Thread Dimitrios Vytiniotis
Hello, Edsko de Vries wrote: > > In this figure, there are rules for annotated abstractions (AABS1 and > AABS2) and annotated terms (ANNOT). What I'm wondering about: are the > rules AABS1 and AABS2 really necessary? As in, if you would remove those > two rules, does there exist a program that ca

[Haskell] Type inference for arbitrary rank types

2006-06-23 Thread Edsko de Vries
Hi, I am currently studying the paper "Practical Type Inference for Arbitrary Rank Types" by Simon Peyton Jones and Mark Shields, and I've been wondering about the final version of the typing rules (figure 7, "Bidirectional version of Odersky-Laufer") (although I suppose the question is slightly m