#9495: Do What I Mean RULES for foldr2 look shady -------------------------------------+------------------------------------- Reporter: dfeuer | Owner: Type: bug | Status: new Priority: highest | Milestone: 7.8.4 Component: | Version: 7.8.3 libraries/base | Keywords: Resolution: | Architecture: Unknown/Multiple Operating System: | Difficulty: Unknown Unknown/Multiple | Blocked By: Type of failure: Runtime | Related Tickets: crash | Test Case: | Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- Comment (by nomeata): The root of all this is the asymmetry of {{{ Prelude> zip [] undefined [] Prelude> zip undefined [] *** Exception: Prelude.undefined }}} So what does the report say? It defined `zip` via `zipWith`: {{{ zipWith :: (a->b->c) -> [a]->[b]->[c] zipWith z (a:as) (b:bs) = z a b : zipWith z as bs zipWith _ _ _ = [] }}} At first I thought that this definition will imply `zipWith (+) [] ⊥ = ⊥`, but that doesn’t seem to be the case – pattern are tried from left to right. So while I dislike the slight asymmetry here, I don’t think it is justified to not follow the standard here. OTOH, I also don’t think that the semantic change, although a wart, is bad enough to justify not being able to fuse on one or the other side. I also think it is reasonable to try to fuse on boths sides (even if the result is not confluent if we could fuse on both sides). So I currently don’t see how to improve the code. Which leaves us improving the documentation. Would you provide a patch against that? -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/9495#comment:3 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler