maintaining pre-AMP+FTP-Prelude in external package

[Henning Thielemann]

On Wed, 13 May 2015, Edward Kmett wrote:

> Keep in mind such a package would have to either supply its own definition for Monad, cutting it off from the
> rest of the world almost entirely and making it near impossible to use with almost any of the existing
> libraries, or users would have to give up defining any new monads.

I am only concerned with exported identifiers. Monad class would remain a 
sub-class of Applicative, but Applicative is not exported by 
PreludePreAMP.