#13021: Inaccessible RHS warning is confusing for users -------------------------------------+------------------------------------- Reporter: mpickering | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.1 Resolution: | Keywords: | PatternMatchWarnings Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by gkaracha): Even though I understand that the message can be confusing, I disagree with calling it in any way "redundant". Pattern matching is side-effectful in a lazy language and in fact we often use it just to force evaluation, so I would not consider it a minor thing. Before GHC 8, both redundant and with-inaccessible-rhs clauses where called "overlapping" which may be more appropriate. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/13021#comment:1 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler

#13021: Inaccessible RHS warning is confusing for users -------------------------------------+------------------------------------- Reporter: mpickering | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.1 Resolution: | Keywords: | PatternMatchWarnings Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by mpickering): How is a user expected to react to this warning? It seems that they can't act on it without changing the semantics of their program. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/13021#comment:3 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler

#13021: Inaccessible RHS warning is confusing for users -------------------------------------+------------------------------------- Reporter: mpickering | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.1 Resolution: | Keywords: | PatternMatchWarnings Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by mpickering): It is easy to trigger cases like this with `COMPLETE` signatures as they are currently designed. For example, {{{ data T = A | B | C {-# COMPLETE B #-} foo :: T -> () foo A = () foo B = () }}} Then the first match is reported as having an inaccessible RHS. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/13021#comment:5 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler

#13021: Inaccessible RHS warning is confusing for users -------------------------------------+------------------------------------- Reporter: mpickering | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.1 Resolution: | Keywords: | PatternMatchWarnings Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by rwbarton): I don't understand what is going on in this example. Suppose `B` was a pattern synonym and we couldn't see its definition. Then the `COMPLETE` pragma could be semantically correct. But if we match `A` before `B` then the match isn't redundant, the behavior of the function could be different in the `A` case. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/13021#comment:7 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler

#13021: Inaccessible RHS warning is confusing for users -------------------------------------+------------------------------------- Reporter: mpickering | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.1 Resolution: | Keywords: | PatternMatchWarnings Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by rwbarton): My confusion here is at a very basic level. `_` is a pattern such that any set of pattern matches which includes `_` is total. Yet in a function like {{{ f :: Maybe a -> Int f Nothing = 0 f _ = 1 }}} the first equation is not redundant in a practical sense, even though the pattern match would still be total without it. There's more to life than totality of pattern matches. I thought that a `{-# COMPLETE p1, ..., pn #-}` pragma was just supposed to mean that a function that matches all of the `pi` should be regarded as total. It doesn't make any sense to me for the compiler to use this knowledge plus the definitions of the `pi` to make any conclusions about whether other values are possible or not. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/13021#comment:9 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler