Repository navigation
Conversation
lehins
commented
Feb 1, 2020
lehins
commented
- Bump up the version in cabal and update changelog.
- Fix markdown header formatting in changelog.
|
Accorfing to PVP
I think we should remove deprecation |
|
I just commented the same thing as @Shimuuar in the related issue. This violates PVP, unfortunately |
* Bump up the version in cabal and update changelog. * Fix markdown header formatting in changelog.
9c1ae17 to
716074a
Compare
|
@chessai I sincerely do not agree with that part of PVP, but at the same time I really don't care about this |
| import qualified Data.List.NonEmpty as NonEmpty | ||
|
|
||
| import qualified Data.Traversable as T (Traversable(mapM)) | ||
|
|
There was a problem hiding this comment.
was this shuffled because of how you wanted to do the cpp to conditionalize mkNoRepType?
There was a problem hiding this comment.
This was shuffled because function definitions should not appear between imports, if I didn't move it it would look like this for base < 4.2.0:
import Data.Data ( Data, DataType, Constr, Fixity(Prefix),
mkDataType, mkConstr, constrIndex )
mkDataType, mkConstr, constrIndex,
mkNorepType )
mkNoRepType :: String -> DataType
mkNoRepType = mkNorepType
import qualified Data.Traversable as T (Traversable(mapM))The slight improvement to the way I did CPP was to reduce multiple imports from the same module, but it had nothing to do with Traversable import. hlint is a great tool that points those flaws to you.
|
I think changing pvp stance on deprecation annotations is ultimately gated on improving how ghc handles those and related warnings. (as with many things, the real way to progress stuff turns into "contrib the nice things you want into ghc and friends") |
|
@lehins this patch looks good, i'd be inclined to undo the shuffle and have two cpp clause so import order styling doesn't get semi spurious reordering. |
|
unrealtedly: |
|
Reverting the ordering of imports is a wrong decision here. See my answer to your comment. |
4 participants
