#7401: Can't derive instance for Eq when datatype has no constructor, while it is trivial do do so. ---------------------------------+------------------------------------------ Reporter: jpbernardy | Owner: monoidal Type: feature request | Status: new Priority: normal | Milestone: 7.8.1 Component: Compiler | Version: 7.6.1 Keywords: deriving | Os: Unknown/Multiple Architecture: Unknown/Multiple | Failure: GHC rejects valid program Difficulty: Unknown | Testcase: Blockedby: | Blocking: Related: | ---------------------------------+------------------------------------------ Comment(by simonpj): Two more thoughts * You point out in #7931, this already works for standalone deriving, which is a bit odd. Standalone and non-standalone deriving should behave pretty much the same. * When I say "works" what I mean is that, for nullary types, all methods `foo` get this definition: {{{ foo = error "Void foo" }}} This is done in `TcGenDeriv.mkRdrFunBind`. Several points: * Is this always the right thing to do? For every method, and every class? Eg for `Read` it wasn't; see #7931. Needs checking. * We could also generate {{{ foo x y = error "Void foo" }}} (ie with the error inside the lambdas) which isn't quite the same. I'm not sure which is right. * Including the name of the type constructor in the error message would be a lot more helpful. Thanks! Simon -- Ticket URL: http://hackage.haskell.org/trac/ghc/ticket/7401#comment:11 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler