[GHC] #13649: RebindableSyntax causes type errors when 'fail' is not defined, even if not used.

[GHC]

#13649: RebindableSyntax causes type errors when 'fail' is not defined, even if not
used.
-------------------------------------+-------------------------------------
        Reporter:  AaronFriel        |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:
       Component:  Compiler          |              Version:  8.2.1-rc1
      Resolution:                    |             Keywords:
                                     |  RebindableSyntax
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  GHC rejects       |  Unknown/Multiple
  valid program                      |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by nomeata):

 This is not really a bug, but rather how `do` notation is specified. If
 you look at
 https://www.haskell.org/onlinereport/haskell2010/haskellch3.html#x8-470003.14
 you will see

 {{{
 do {p <- e; stmts} = let ok p = do {stmts}
                          ok _ = fail "..."
                      in e >>= ok
 }}}

 One could argue that if the pattern is `_`, the second line should be
 omitted, but where to stop? What about `do () <- …`, an obviously complete
 pattern, should the `fail` line be omitted as well? But that information
 might only be available after type-checking, whereas this happens before…

 I don’t have an opinion of my own here, just pointing out that the current
 behaviour is as specified, and that a change is not as trivial as it
 looks.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13649#comment:1>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler