On Tue, 06 Jun 2000, Koen Claessen wrote:
> Now, obviously there are not that many people who use
> polymorphic let-bindings in do, but I don't think that is a
> reason to introduce this restriction.
I agree. We were just trying to see if such polymorphic
let-generators were used often in practice.
> On the other hand, recursive do-notation is obviously is a
> great feature that many (at least I) have been wanting for
> some time now. If that feature implies monomorphic let
> bindings in do-notation, then there is clearly a trade-off
> to make. But don't use the argument "nobody is using this
> feature" to make the language design less consistent.
>
> Now, send us all a pointer to your work! :-) (I want to read
> about it!!),
Here's the pointer:
http://www.cse.ogi.edu/PacSoft/projects/muHugs/
The paper titled "Recursive Monadic Bindings" explains the basics. (It also
briefly talks about the let-problem.)
The problem with let-generators are detailed in the paper titled:
"A Recursive do for Haskell: Design And Implementation"
