[m-dev.] Re: For review: changes to graph.m

[Fergus Henderson]

David Glen JEFFERY, you wrote:
> 
> > Given this argument, and given that a predicate written for a type class
> > doesn't use operations not promised by that type class, then I'd argue
> > that different types in a type class must have the same semantics as far
> > as can be seen from that type class.
> 
> One of the classic examples of the problems caused by overlapping instances is
> something like:
> 
> typeclass Showable a where ...
> 
> In this example, there are two instances which have different semantics. They
> are also overlapping, and using the wrong one would be a Bad Thing(TM).

That's actually a different problem.  Gofer _does_ allow that, I think.

Peter was talking about allowing the system to pick a type for you,
any type which satisfied the constraints.  (That's what I mistakenly
said that I thought gofer did, in my original mail.)
Here's an example of why letting the system pick a type for you
causes problems:

	main --> read(X), write(X).

Anything which is a member of the type classes `readable' and `writeable'
should work, right?  But clearly this will have different effects depending
on which type is chosen.

-- 
Fergus Henderson <fjh at cs.mu.oz.au>   |  "I have always known that the pursuit
WWW: <http://www.cs.mu.oz.au/~fjh>   |  of excellence is a lethal habit"
PGP: finger fjh at 128.250.37.3         |     -- the last words of T. S. Garp.