On Thu, 2007-03-22 at 14:50 +0000, Neil Mitchell wrote:
Hi
Note that in the bytestring case the compiler doesn't do so, it's the library that does (by telling the compiler to rewrite things with RULEs).
ByteString is in base, and ships with GHC, I think that makes it GHC's fault if it does naughty things.
Unfortunately, it's very appealing in bytestring's case...
Yes, the dilema between performance and purity...
It's not impure. Our rewrite rules for ByteStrings can refine your code (in the domain / CPO sense) by making it lazier. It doesn't 'invent' random values to replace _|_ it gives you the result you'd have got if the data type were slightly lazier. The optimised version doesn't force things that are not necessary for calculating the result and if any of those things you would have forced were _|_ then we've done a refinement rather than a strict equality rewrite. We never rewrite in the opposite direction since that'd be bad. Duncan