[GHC] #1308: Type signature in warning is wrong

[GHC]

#1308: Type signature in warning is wrong
-------------------------+--------------------------------------------------
    Reporter:  guest     |        Owner:         
        Type:  bug       |       Status:  new    
    Priority:  normal    |    Milestone:         
   Component:  Compiler  |      Version:  6.7    
    Severity:  normal    |   Resolution:         
    Keywords:            |   Difficulty:  Unknown
          Os:  Windows   |     Testcase:         
Architecture:  Unknown   |  
-------------------------+--------------------------------------------------
Comment (by Isaac Dupree):

 Given inferred monomorphic type `Inferred type: autoChart :: t View`,

 I think it would be desirable to (additionally) report the type as if the
 M-R didn't apply, to be informative and in case this is the signature that
 the user wants to add.

 Also it would be helpful to report the actual type GHC decides upon
 (unless there's a type error?). I assume that GHC knows what 't' becomes
 after applying M-R.

 Now, is it useful to report `autoChart :: monomorphic t. t View`?  As far
 as I know, it is easy to see what this could be, just based on the fully
 polymorphic type,  especially if a specific example - the actual
 monomorphic type GHC has chosen - is shown as well.

 Hmm... is it even possible to pretend that one thing (e.g. `autochart`) is
 polymorphic (as suggested above), in the presence of other things falling
 under the monomorphism restriction?  Or are there some cases where it
 still might not be clear what to report?

-- 
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/1308>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler
-------------- next part --------------
_______________________________________________
Glasgow-haskell-bugs mailing list
Glasgow-haskell-bugs at haskell.org
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs