- User Since
- Dec 14 2015, 9:33 PM (208 w, 1 d)
Nov 24 2016
Apr 4 2016
Merged the all "Identity and Const" sections of the changelog.
Apr 3 2016
Fixed previous diff where I accidentally enabled DeriveDataTypeable unnecessarily in Data.Functor.Const
Apr 2 2016
Dec 21 2015
I don't know why this patch is being rejected by Harbormaster. It's working on my clone. Maybe there's something I'm missing. If somebody else could have a look at this for me I would appreciate it.
Rebased on top of git master, hopefully this will fix the build errors
Dec 19 2015
The build failed because the updates to base.cabal from D1543 conflicted with this. I've rebased on top of the current git master branch and also fixed that typo in the changelog.
Removed erroneous @since annotation on Const data type
Dec 15 2015
Forgot to update base.cabal
I added a version of this patch that's rebased on top of D1630.
I accidentally mixed this up with the changes I'm making to Const in D1626. This keeps them orthogonal.
Or I could just make a patch that implements that (moving Const out of Control.Applicative) (but re-exporting it obviously), and then if people want to make WrappedMonad kind-polymorphic later they can do that then.
Do we want WrappedArrow and WrappedMonad to be kind-polymorphic as well? Should I change change the documentation or should I add kind signatures to WrappedMonad and WrappedArrow restricting them to *? Especially seeing as there is talk of splitting Const off into its own module Data.Functor.Const as well.
With the last change, it does indeed ./validate now.
It turns out Const already had a Semigroup instance defined in Data.Semigroup. Ix should have been imported from GHC.Arr instead of Data.Ix to avoid circular dependencies.
Dec 14 2015
Added changelog entry