Curious behaviour of irrefutable pattern.

[Bernie Pope]

On 22/12/2006, at 7:14 PM, Simon Peyton-Jones wrote:

> | I take it, then, that the answer to the question of "under what
> | circumstances does the (error -> non-termination) transformation
> | happen?" is that GHC can choose among different bottoms that are
> | present in the program.  It can't, however, willy-nilly convert my
> | error calls to bottom.  (Or something more precise along the same
> | lines.)
>
> Yes, that's right, good point.

I've been following this thread with interest.
But now I don't understand what the conclusion is.

I thought Simon's earlier example showed that the semantics of
GHC does not distinguish between an infinite loop and a call to error.
Otherwise the strictness optimisation would be unsound.

Mike, what do you mean by "willy nilly convert my error calls to  
bottom"?

Cheers,
Bernie.