I'm pleased to note that thanks to Malcolm Wallace, the patch has been pushed: http://hackage.haskell.org/trac/ghc/ticket/3292#comment:7 Y'all can look forward to a 'void' in GHC 6.14, likely. -- gwern