[GHC] #3851: type family does not get expanded in ghc-6.12.1

GHC trac at galois.com
Mon Feb 8 11:33:22 EST 2010


#3851: type family does not get expanded in ghc-6.12.1
----------------------------------------+-----------------------------------
    Reporter:  kosmikus                 |        Owner:                           
        Type:  bug                      |       Status:  new                      
    Priority:  normal                   |    Milestone:                           
   Component:  Compiler (Type checker)  |      Version:  6.12.1                   
    Keywords:                           |   Difficulty:                           
          Os:  Unknown/Multiple         |     Testcase:                           
Architecture:  Unknown/Multiple         |      Failure:  GHC rejects valid program
----------------------------------------+-----------------------------------

Comment(by simonpj):

 It's because of an awkward interaction.  When matching a GADT, GHC
 currently requires the context of the pattern to be "rigid", so that it
 knows exactly what type refinement happens as a result of the match.
 Well, is a type-function application rigid?  Presumably not in 6.12.  I'm
 not sure why it was considered rigid in 6.10.

 All this grunge about rigidity will be swept away with the new outside-in
 type inference machinery (see our ICFP'09 paper), so I'm quite reluctant
 to try to "fix" this.  Can you work around it for now, perhaps by using
 less perspicuous type signatures?

 Simon

-- 
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/3851#comment:1>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler


More information about the Glasgow-haskell-bugs mailing list