
#14558: Unable to parse integer-gmp's Cabal file -------------------------------------+------------------------------------- Reporter: taylorfausak | Owner: hvr Type: task | Status: new Priority: normal | Milestone: Component: Core Libraries | Version: 8.2.2 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by snoyberg):
I think the former can be arranged, but then it's off-topic for the GHC bug tracker.
Where would be the right place to file issues like this? Since integer-gmp is part of GHC, I would have guessed that this is the best place to report. Perhaps the integer-gmp.cabal file needs to be modified to include an issue tracker URL?
And as Phyx mentioned above, GHC 8.2 is not a part of any LTS release, so IMO that statement of mine remains correct.
I disagree; Stackage Nightly is not intended to have the same level of stability as an LTS release, but: 1. We strive for a high level of stability, and 2. In this case, a cabal file was uploaded to Hackage months after the release of GHC, which caused problems down the line. I'd like to understand the process that allowed this situation to occur, and ideally ensure that it doesn't happen in the future. We're not that far away from an LTS release including GHC 8.2. If the upload occurred in February instead, we may have been facing this bug with an LTS release. To quote myself from above:
Can you clarify how this has been resolved? The file on Hackage and in 01-index.tar.gz both appear to still have cabal-version: 2.0, preventing Stack 1.5 from building with GHC 8.2. Also, it would be helpful to know what steps have been taken to ensure it won't happen in the future, and what led to this situation in the first place.
There's still not been any real answer to any of these points; Herbert simply stated that the issue has resolved and won't occur in the future. We're assuming that by resolved he means that the file was updated in the Git repo (though it doesn't address the tagged release), and still don't know what "we've taken steps that will ensure this won't be an issue in the future anymore" means. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/14558#comment:39 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler