I've finally got around to looking at this. Turned out to
be that the type/class dependency analysis wasn't expecting types
in the context of a class. But you had
class Foo T a => Baz a where ...
(legal in GHC-extensions). Easily fixed, and done in the 4.08 branch.
Thanks for reporting it. If you check our the 4.08 head you should
find both this and your other bug are dealt with --- without introducing
lots of new functionality!
Simon
| -----Original Message-----
| From: George Russell [mailto:[EMAIL PROTECTED]]
| Sent: 01 November 2000 17:24
| To: [EMAIL PROTECTED]
| Cc: Simon Marlow
| Subject: Why is variable out of scope?
|
|
| (cc:d to Simon Marlow since my last bug report doesn't seem
| to have got through)
|
| When I attempt to compile the attached file with (a) with
| ghc-4.08/Solaris and
| (b) ghc-4.08.1/Linux I get the following message:
|
| TestSh.hs:8:
| `Shape' is not in scope
| When checking kinds in `HasConfigValue Shape nodeTypeParms'
| In the class declaration for `HasShape'
|
| Compilation had errors
|
| (command line :
| ghc TestSh.hs -c -syslib lang -fglasgow-exts
| )
|
| So how on earth is Shape not in scope, when I defined it two
| lines previously?
|
_______________________________________________
Glasgow-haskell-bugs mailing list
[EMAIL PROTECTED]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs
