> On Fri, Jun 06, 2003 at 09:32:18AM +0100, Simon Peyton-Jones wrote: > > Yes! Yes! Advice is good! > > OK, how about "avoid unsafePerformIO like the plague"? > > Why is it the business of the FFI spec to document unsafe uses of > unsafePerformIO? I'd like to second Ross here. Advice is good at the right place at the right time. Cheers, Ralf