Changelog of @hackage/some 1.0.5

1.0.5

  • Add EqP and OrdP classes. These are strong versions of Eq1 and Ord1, and on the other hand weaker versions of GEq and GCompare. They are exactly what's needed for Eq and Ord instances of Some.

    The naming is unfortunate: GShow would be better named ShowP, as it's similar version of Show1.

    Note: we could add ReadP with readsPrecP :: Int -> ReadS (t a) method, but it will barely have any instances. GRead is different, as it can reify the type index for many types, e.g. for the singletons.

    In some future there will be major version of some with following breaking changes:

    • EqP and OrdP will become superclasses of GEq and GCompare
    • Eq (Some t) will require EqP t, similarly for Ord and OrdP.
    • GShow will get forall a. Show (f a) superclass. (This will cause removal of Product and Sum instances for base <4.18).

    To ease future transition you may

    • Define EqP and OrdP instances for your types. The defaultEq and defaultCompare methods can be used to define eqp and comparep from GEq and GCompare instances respectively.
    • Move to use GHC.Generics.:*: and :+: instead of Data.Functor.Product and Sum, as these have better Eq and Ord instances.

1.0.4.1

  • Drop support for GHC before 8.6

1.0.4

  • Add instances for (:~~:)
  • Add instances for :+: and :*:
  • Add defaultGeq :: GCompare f => f a -> f b -> Maybe (a :~: b)
  • Add defaultGshowsPrec :: Show (t a) => Int -> t a -> ShowS

1.0.3

  • Make GNFData PolyKinded.
  • Add GNFData ((:~:) a) and GNFData TypeRep instances

1.0.2

  • Explicitly mark Data.Some as Safe. It was previously inferred, yet it was Safe too, as it only re-exports other explicitly marked modules.
  • Allow base-4.15, GHC-9.0 compatibility

1.0.1

1.0.0.3

  • One less unsafeCoerce (thanks to David Feuer)

1.0.0.2

  • Broken release

1.0.0.1

1

  • Split out of dependent-sum
  • Have GADT, Newtype, Church variants
  • Add NFData instance