[GUI] Re: GUI Library Task Force

[Manuel M. T. Chakravarty]

[redirected to the GUI mailing list]

Matt Harden <matth@mindspring.com> wrote,

> "Manuel M. T. Chakravarty" wrote:
> 
> > * I am not a big fan of introducing an extra monad (`GUI' in
> >   this case).  It can easily become a pain in programs that
> >   do a lot of "normal" IO as you have to lift all IO
> >   functions to GUI.
> 
> Shouldn't GUI be a typeclass (as a subclass of Monad), with IO being an
> instance of it?  This gives the following benefits:
> 
> 	o  "Pure" GUI functions can reflect this in their type.
> 		i.e. (GUI g) => g ()
> 	o  IO functions can use GUI features directly without lifting.
> 	o  "Pure" GUI functions are implicitly usable in the IO monad
> 		They "are" IO functions.

That's an interesting idea.  The main problem that I am
seeing at the moment is that the basic GUI functionality
that usually is imported from C libraries via the FFI will
be in IO.  That's because foreign imports of sideeffecting
functions (which almost all basic GUI functions are) are
required to be of type `a -> IO b'.  This constraints the
GUI monad to be IO or an extension of IO; unless you want to
get into some rather ugly messing around with `unsafePerformIO'.

Cheers,
Manuel