Re: [Haskell-cafe] Parallel weirdness [new insights]