Changelog for the ghc-typelits-knownnat
package
- Support for GHC 9.10.1
- Fix infinite loop between plugin and solver pipeline
- Work around GHC issue 23109
- Support for GHC 9.8.1
- Support for GHC-9.6.0.20230210
- Add support for GHC 9.4
- Add support for GHC 9.2.0.20210422
- Raise upper limit for TH dep to allow building on ghc-9.0.1
- Add support for GHC 9.0.1-rc1
- Add support for GHC 8.10.0-alpha2
- Solve "known" type-level Booleans, also inside
If
(GHC 8.6+)
- Move
KnownNat2
instances forDiv
andMod
fromghc-typelits-extra
toghc-typelits-knownnat
- Fix Inferred constraint is too strong #19
- Add support for GHC 8.5.20180306
- Add support for GHC 8.4.1
- Add partial GHC 8.4.1-alpha1 support
- Drop
singletons
dependency #15KnownNatN
classes no longer have theKnownNatFN
associated type family
- Fix testsuite for GHC 8.2.1
- GHC 8.2.1 support: Underlying representation for
KnownNat
in GHC 8.2 isNatural
, meaning users of this plugin will need to update their code to useNatural
for GHC 8.2 as well.
- New features:
- Solve normalised literal constraints, i.e.:
KnownNat (((addrSize + 1) - (addrSize - 1))) ~ KnownNat 2
- New features:
- Derive smaller constraints from larger constraints when they differ by a single variable, i.e.
KnownNat (a + b), KnownNat b
impliesKnownNat a
.
- Derive smaller constraints from larger constraints when they differ by a single variable, i.e.
- Fixes bugs:
- Source location of derived wanted constraints is, erroneously, always set to line 1, column 1
- New features:
- Handle
GHC.TypeLits.-
- Handle custom, user-defined, type-level operations
- Thanks to Gabor Greif (@ggreif): derive smaller from larger constraints, i.e.
KnownNat (n+1)
impliesKnownNat n
- Handle
- New features: Solve "complex" KnownNat constraints involving arbitrary type-functions, as long as there is a given KnownNat constraint for this type functions.
- Fixes bug: panic on a non-given KnownNat constraint variable
- Initial release