Hi Lennart,

It would be nice if HLint didn't suggest things that it will object to in the next round. Like

LLVM/Core/CodeGen.hs:176:1: Eta reduce Found: applyArgs f g = apArgs 0 f g Why not: applyArgs f = apArgs 0 f

The idea is to specify things one step at a time, so the user learns from the hints, as well as fixing them. However, in the case of eta reduction, doing several reductions at a time does make sense. I'll add it to the todo list. Thanks Neil