Hi, Am Freitag, den 25.04.2008, 11:49 -0700 schrieb Dan Weston:
I guess like minds think alike! See the very recent e-mail thread started by Ryan Ingram: http://thread.gmane.org/gmane.comp.lang.haskell.cafe/39155/focus=39159
Take a look at the code referenced in Luke Palmer's reply: http://luqui.org/git/?p=luqui-misc.git;a=blob;f=work/code/haskell/frp/Fregl/...
A snippet follows:
class (Monad m) => MonadSuspend v m | m -> v where attempt :: m a -> m (Either a (v -> m a)) suspend :: m v
newtype SuspendT v m a = SuspendT { runSuspendT :: m (Either a (v -> SuspendT v m a)) }
Your (Coroutine m a) appears to be isomorphic to (SuspendT () m a) [so Coroutine m () = SuspendT () m ()]
Your runCoroutineT appears to be isomorphic to a specialization of runSuspendT:
runSuspendT' :: SuspendT () m () -> m (Either () (() -> SuspendT () m ()))
Here the () -> a ~ a and Either () a ~ Maybe a
You are quite right, it really is the same thing. The implementation behind runCoroutineT is not just a specialization, but the exact same thing (with Left and Right switched). I just put the specialization there because I had no need for a return value in my use case. And interesting how Ryan and me had the same thoughts on the same day. Maybe the April 24th should be considered Suspend You Monadic Action Day. Greetings, Joachim -- Joachim Breitner e-Mail: mail@joachim-breitner.de Homepage: http://www.joachim-breitner.de ICQ#: 74513189 Jabber-ID: nomeata@joachim-breitner.de