[GHC] #6074: Instance inference failure with GADTs
GHC
cvs-ghc at haskell.org
Fri May 4 15:20:39 CEST 2012
#6074: Instance inference failure with GADTs
---------------------------------+------------------------------------------
Reporter: goldfire | Owner:
Type: feature request | Status: new
Priority: normal | Milestone:
Component: Compiler | Version: 7.5
Keywords: GADTs | Os: Unknown/Multiple
Architecture: Unknown/Multiple | Failure: None/Unknown
Difficulty: Unknown | Testcase:
Blockedby: | Blocking:
Related: |
---------------------------------+------------------------------------------
Changes (by goldfire):
* type: bug => feature request
Comment:
Yes, I was worried about all of these problems when I made the post. It
seems possible to circumvent the problem michalt brings up with some
strictness guarantee to eliminate the possibility of {{{undefined}}} and
friends. And, in truth, my situation is the one dreixel points out, where
I'm using a closed data kind.
But, of course, simonpj's point is well taken and was my chief worry. I
guess my best hope would be that there could be some construct that
produces the pattern-match automatically when given the term to match over
and the branch expression. In fact, I played around writing some such
construct in Template Haskell, but TH doesn't have access to inferred
types of expressions, so it's quite clunky to use.
I doubt there's much interest in this particular feature, so I'll just
change the ticket to be a feature request and go back to the drawing board
on my own design. Thanks for the feedback.
--
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/6074#comment:4>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler
More information about the Glasgow-haskell-bugs
mailing list