It's too bad that indexes are `Int` instead of `Word` under the hood. Why is `Int` used in so many places where it is semantically wrong? Not just here but also in list indexing... Indices/offsets can only be positive and I can't see any good reason to waste half the address space -- yet we encounter this problem over and over again.
Readers who disliked the above also disliked the following: "index out of range" error message regression http://hackage.haskell.org/trac/ghc/ticket/2669 Int / Word / IntN / WordN are unequally optimized http://hackage.haskell.org/trac/ghc/ticket/3055 Arrays allow out-of-bounds indexes http://hackage.haskell.org/trac/ghc/ticket/2120 .. Not to mention that many serious array programmers use their own array libraries (yes, plural:-(, bypassing the "standard", so their valuable experience/expertise doesn't result in improvements in the standard array libraries (nor have they agreed on a new API). If any of this is affecting your use of GHC or libraries, you might want to add yourself to relevant tickets, or add new tickets. Small bug fixes, alternative designs and grand array library reunification initiatives might also be welcome. Claus PS. You could, of course, rebase your array indices to make use of the negatives, so the address space isn't wasted, just made difficult to use.