Changelog for barbies-1.1.3.0
Changelog for barbies
1.1.3.0
Wear
will raise aTypeError
instead of getting stuck (Alex Peitsinis).
1.1.2.1
- Uploaded 1.1.2.0 was broken (missing
btraverseC
)
1.1.2.0
- Add
traverseC
(Ole Krüger). - Fix typo in ProductB laws (thanks to Ben Radford).
1.1.1.0
- Add
bmapC
(Chris Penner).
1.1.0.0
-
Make all classes poly-kinded (#7): a barbie can now be any type parameterised by a type
(k -> Type)
. In particular, a (higher-kinded) barbie is a type parameterised by a barbie. Thanks to Ole Krüger. -
Add instances for functor transformers:
Proxy
,Const
,Product
,Sum
andCompose
(Ole Krüger).
1.0.0.0
-
Replaced
ConstraintsOf
inConstraintsB
byAllB
, which allows constraints to be given ona
instead of onf a
. TheClassF
class lets us specify constraints onf a
by doingAllB (ClassF c f) b
.ConstraintsOf
becomes then a type alias. Credit goes to Csongor Kiss. -
ConstraintsOf
was ultimately deprecated in favour ofAllBF
, which is shorter and more consistent withAllB
. -
Renamed
ConstraintsB(adjProof)
toConstraintsB(baddDicts)
. -
Renamed
ProofB(bproof)
toProductBC(bdicts)
. -
Changed the way
Wear
works: now wear-types need to have an extra type parameter that controls whether they areBare
orCovered
. This let us remove all the "magic" that was involved, in the sense that one couldn't have instances ofFunctorB
, etc, for wear-types wihtout usingunsafeCoerce
(this was true also for handwritten instances). -
Added
bsequence'
, a frequent specialisation ofbsequence
. -
Added
bfoldMap
. -
Added
buniqC
andbmempty
. -
Improved the internal instance derivation mechanism. We no longer need
unsafeCoerce
and the code should be in general indistinguishible from hand-written instances (not currently verified). -
Fixed support for barbie-types that have additional type parameters (#5).
0.1.4.0
-
Added
btraverse_
-
Added the trivial
Void
andUnit
barbies
0.1.3.1
- Fixed issue on Barbie-types with strictness annotations.
0.1.3.0
- Use both
Monoid
andSemigroup
as constraints for theMonoid
instance, so that this works with ghc 8.0 to 8.4 (Fraser Murray)
0.1.2.0
- Use
Monoid
and notSemigroup
as constraints for theMonoid
instance
0.1.1.0
- Added
instance Semigroup (Barbie b)
to go along theMonoid
instance
0.1.0.1
- Works under GHC 8.0.2, but notice one needs to use empty instance
declarations, because ghc chokes on
deriving
clauses.