Re: [Haskell-cafe] Fwd: [Haskell-beginners] Monad instances and type synonyms

The point in not allowing partially applied type synonym instances is that it'd make deciding whether a type is an instance of a class much harder. Cf. here[1] for a similar question with the Category class. -- Steffen [1] Attached message. Couldn't find it on the archives.. On 04/14/2013 07:10 AM, Christopher Howard wrote:

I asked this question in Haskell-beginners, but I haven't heard anything yet, so I'm forwarding to Cafe.

-------- Original Message -------- Subject: [Haskell-beginners] Monad instances and type synonyms Date: Sat, 13 Apr 2013 17:03:57 -0800 From: Christopher Howard Reply-To: The Haskell-Beginners Mailing List - Discussion of primarily beginner-level topics related to Haskell To: Haskell Beginners

I am playing around with some trivial code (for learning purposes) I wanted to take

code: -------- -- SaleVariables a concrete type defined early

-- `Adjustment' represents adjustment in a price calculation -- Allows functions of type (a -> Adjustment a) to be composed -- with an appropriate composition function type Adjustment a = SaleVariables -> a --------

And put it into

code: -------- instance Monad Adjustment where

(>>=) = ... return = ... --------

If I try this, I get

code: -------- Type synonym `Adjustment' should have 1 argument, but has been given none In the instance declaration for `Monad Adjustment' --------

But if I give an argument, then it doesn't compile either (it becomes a "*" kind). And I didn't want to make the type with a regular "data" declaration either, because then I have to give it a constructor, which doesn't fit with what I want the type to do.

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe