> I can't imagine a system in which nil ISN'T falsy... In the HOPL paper we say that, if things were different, using nil in a logical expression should probably raise an error, just like in arithmetic expressions.