[Haskell-cafe] Re: Hackage Build Failures

[Ben Franksen]

Emil Axelsson wrote:
> Stephan Friedrichs skrev:
>> Emil Axelsson wrote:
>>> [...]
>>>
>>> The error comes from using QuickCheck 2, which happens to also use the
>>> operator (><). I can see two ways to solve the problem:
>>>
>>> (1) Add "< 2" after "QuickCheck" in the Wired.cabal file.
>>>
>>> (2) Add "hiding ((><))" after "import Test.QuickCheck" in
>>> Data/Hardware/Internal.hs
>> 
>> Emil,
>> 
>> my suggestion is: Please use alternative (2), at least if there are no
>> further problems with quickcheck 2! Otherwise, depending on quickcheck
>> <2 just introduces unnecessary package incompatibilities.
> 
> Hi Stephan,
> 
> Option (1) was only meant as a temporary solution for Cetin. It just
> occurred that the obvious solution to make both QC1 and QC2 users happy is
> to have an explicit import list rather than hiding (><) (there are no
> further problems with QC2). I'll do this until QC2 is standard in GHC.

I stumbled over the same thing in a project of mine. I must say that I find
it very unfortunate that QuickCheck-2 exports this operator, since it
clashes with imports of Data.Sequence. Could the decision to use this
operator in QuickCheck be revised (and similarly export of .&. which
clashes with the one in Data.Bits)?

I think these operators should have a 'standard' meaning.

Cheers
Ben