comonads, io

[Ashley Yakeley]

In article <20030102225641.GA24187@smtp.alicorna.com>,
 Andrew J Bromage <ajb@spamcop.net> wrote:

> 	comain :: OI a -> ()
> 	comain w
> 	    = coeval (w .>> show (a,b) =>> stdPutStrLn)
> 	    where
> 		a = coeval (w .>> () =>> stdGetChar)
> 		b = coeval (w .>> () =>> stdGetChar)
> 
> Even though a and b are identical, they return different values.  I see
> two possibilities: Either my intuition is way off, or the OI comonad
> breaks referential transparency.

So is Kieburtz smoking crack, or are we writing OI-style programs 
incorrectly? One possibility is that comonads are useful for some 
things, but not for doing IO actions safely.

-- 
Ashley Yakeley, Seattle WA