Re: More questions

[kh]

> More questions along the same lines as for n+k:
> 
> Does == in the pattern match translation refer to == in PreludeCore?
> 
> Does negate in the translation of -e refer to negate in PreludeCore?

All identifiers used in explicit translations refer to those 
from the Prelude.

Kevin

More questions

[Lennart Augustsson]

More questions along the same lines as for n+k:

Does == in the pattern match translation refer to == in PreludeCore?

Does negate in the translation of -e refer to negate in PreludeCore?

-- Lennart

Re: More questions

[jhf]

| More questions along the same lines as for n+k:
|
| Does == in the pattern match translation refer to == in PreludeCore?
|
| Does negate in the translation of -e refer to negate in PreludeCore?
|
| -- Lennart

I recall that we did decide that such references to standard functions
in the report are always to be understood to be the Standard Prelude
functions, even when the names may have been locally rebound.  I don't
find a clear statement to this effect in the report, though.  (Maybe
I didn't look hard enough.)

Note, however, that even under this interpretation, these syntactic
translations may still involve code supplied by the user, but such
code can only be instance declarations for user-defined types in
the standard classes, not some arbitrary function binding that may
not even be type-correct in the translation.

--Joe

More questions on hiding.

[haskell-request]

Original-Via: uk.ac.nsf; Wed, 23 Oct 91 23:39:10 BST

What happens is you import a module, A, and it contains a value, say
x::T
and also a type synonym for T, say
type T = Int
but then you hide T on import. I.e.
import A hiding(T)
The type synonym T is not imported any other way.  Is this an error
or is there a special case for type synonyms (like expanding them).
My specific worry about this is that you can do
import Prelude hiding(ShowS)
thereby rendering the Text instance useless (even wrong if you
use anything in it).  The problem is that the type synonyms
ReadS and ShowS are in Prelude and not PreludeCore.  What I really
hope is that this is a mistake and that they should be in PreludeCore.

In hope of an answer from the great and knowledgable Haskellites
-- Lennart