[GHC] #8539: Data.Complex shouldn't use default implementation of (**)

[GHC]

#8539: Data.Complex shouldn't use default implementation of (**)
-------------------------------------+-------------------------------------
        Reporter:  jjaredsimpson     |            Owner:
            Type:  bug               |           Status:  patch
        Priority:  low               |        Milestone:
       Component:  Prelude           |          Version:  7.6.3
      Resolution:                    |         Keywords:
Operating System:  Unknown/Multiple  |     Architecture:  Unknown/Multiple
 Type of failure:  Incorrect result  |       Difficulty:  Easy (less than 1
  at runtime                         |  hour)
       Test Case:                    |       Blocked By:
        Blocking:                    |  Related Tickets:
-------------------------------------+-------------------------------------

Comment (by Scott Turner):

 > In my opinion complex numbers with zero imaginary part should behave
 like simple real numbers in all such functions.
 This is a naive understanding of the complex numbers. They are not a plain
 extension of the real numbers; they are missing the linear ordering. So
 you can't say that log(0)= -infinity as you do with the real numbers,
 because there are additional ways in which to approach 0.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8539#comment:13>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler