[Haskell] FRP and Yampa

[Brett G. B. Wortzman]

Hi all,

I'm doing my undergraduate thesis on monads, comonads, and arrows and
their applications.  I've been focusing part of my research on the Yale
Haskell group's work on Functional reactive programming, and specifically
on their move from a monadic to an arrow-based implemenation (as in
http://haskell.cs.yale.edu/yale/papers/haskellworkshop02/index.html).
I've been unable to find any literature that really gives a good
explanation of why they chose to make this move, however.  Can anybody
shed some light on why arrows were seen as a better choice for this?

Also, does anyone know if there's been any research about implementing FRP
with comonads?  I can't find any literature about it, but I'm not sure if
it's never been looked into, or just attempted and determined to be a bad
idea.  If the latter, I'd be interested to know why.

Thanks,
Brett Wortzman