Re: [Haskell-cafe] How is laziness defined?

G'day all. Quoting Matthew Brecknell :

In other words, the irrefutability of a pattern match does not distribute inside the top-level data constructor of the pattern.

I wasn't disagreeing with you, which is why I didn't comment. Note also that if Haskell prime incorporates bang patterns, and hence "strict lets", this will make the situation even more subtle. Cheers, Andrew Bromage