Sweet! Thank you very much! Just out of curiosity: how does this differ from the following, not compiling type signature?

library :: forall t. NetworkDescription t (Behavior t String) -> IO ()

Regards, Alexander Foremny 2012/6/27 Sjoerd Visscher :

This should work:

   library :: (forall t. NetworkDescription t (Behavior t [Char])) -> IO ()

greetings, Sjoerd

On Jun 27, 2012, at 4:39 PM, Alexander Foremny wrote:

...

Hello list,

I am currently experimenting with Reactive.Banana and I am trying to write a program that simply displays a String which is supplied by the user. Thus in Main.hs I'd like to create a Behavior t String from some arbitrary sources and pass this to a library function performing the printing. However, I am not able to get the desired behavior of my program. A sample Main.hs is available as Gist [1].

The cause of error appears to be that the type t over which NetworkDescription and Behavior are parameterized are distinct! I tried several variations of the code in [1] using ExistentialQuantification over t and/or giving explicit type signatures. The error is slightly different in each case, but always boils down to the same reason.

In case explicit error messages are required I can post them to the list.

Any insight into this would be of great help. And additionally I'd be interested if this kind of program is possible with Reactive.Banana at all.

[1] https://gist.github.com/3004430

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

-- Sjoerd Visscher https://github.com/sjoerdvisscher/blog

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

On 6/27/12 2:57 PM, Alexander Foremny wrote:

Sweet! Thank you very much!

Just out of curiosity: how does this differ from the following, not compiling type signature?

...

library :: forall t. NetworkDescription t (Behavior t String) -> IO ()

The type: (forall a. F a) -> G is isomorphic to: exists a. (F a -> G) which differs in the obvious way from: forall a. (F a -> G) Or, from a game-theoretic perspective: with the first type it's the callee that gets to decide which type A is; so the caller only knows that such an A exists, but they have no knowledge or control over what it is. Whereas with the latter type, it's the caller that gets to choose A, and the callee has to deal with it no matter which type it is. -- Live well, ~wren