Re: Over loaded strings test case

Lennart Augustsson <[email protected]> Tue, 23 Jan 2007 08:54:55 -0500
Newsgroups gmane.comp.lang.haskell.cvs.all,gmane.comp.lang.haskell.cvs.ghc
Message-ID <[email protected]>
Yes, there's a problem with default.  Simon PJ said he would fix
defaulting, but I guess he's not done that yet.  Ask him, and if
he's busy I can take a look again.  The defaulting code is a bit
convoluted. :)

	-- Lennart

On Jan 23, 2007, at 06:58 , Simon Marlow wrote:

> Ian Lynagh wrote:
>> Hi Lennart,
>> I've applied all the overloaded-strings patches. Thanks!
>
> Looks like there's one failure in nofib/spectral/power related to  
> this:
>
> ==nofib== power: time to compile Main follows...
> /playpen/ghc/nightly/HEAD-cam-02-unx/i386-unknown-linux/compiler/ 
> stage3/ghc-inplace -H16m -O -O -Rghc-timing -H32m -hisuf hi -c  
> Main.hs -o Main.o
>
> Main.hs:15:0:
>     No instances for (IsString Double,
> 		      IsString Rational,
> 		      IsString Integer)
>       arising from a 'default' declaration at Main.hs:15:0-34
>     Possible fix:
>       add an instance declaration for
>       (IsString Double, IsString Rational, IsString Integer)
>     when checking that each type in a default declaration
>     is an instance of class Num
>
> Cheers,
> 	Simon