ANN: H98 FFI Addendum 1.0, Release Candidate 10

[Alastair Reid]

On Monday 02 June 2003 2:32 pm, Simon Marlow wrote:

> Ok.  But I still don't understand why the whole discussion isn't moot.
> I can't see how to acquire a value of type T that isn't bottom.  

Whether you can acquire values of this type or not, we need to give it a 
semantics.

We know that T must contain bottom.

I'd argue that even if we can't manipulate or observe them, the
 semantics should admit that the type contains a bunch of other
 (foreign) values because it is more accurate and having an
 accurate semantics is bound to pay off.

> Could you give an example?

No , I probably can't come up with an example as things stand at the moment.  
But who knows what changes we might make in the future and when we do, we're 
bound to do better if our semantics relfects reality instead of relying on a 
trick.

--
Alastair