You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'd like to add TemplateHaskell to support automatically deriving Hashable instances for users' datatypes. These should replicate instances for Either, tuples, (), Maybe, but not necessarly [] ( as I haven't quite figured out the rule there).
The only slight trickiness is we want to make sure to only constrain the deriving type's parameters to Hashable when it appears in the RHS (i.e. phantom type parameters should not be constrained to Hashable).
I don't think it's worth using GHC.Generics since it's nearly as awful to use and probably much slower.
The text was updated successfully, but these errors were encountered:
Although it's probably a requirement that I implement a drop-in replacement to hashable's auto-deriving with GHC.Generics if I want this library to have any traction (that way e.g. unordered-containers could switch to this without too much breakage.).
I'd like to add
TemplateHaskell
to support automatically derivingHashable
instances for users' datatypes. These should replicate instances forEither
, tuples,()
,Maybe
, but not necessarly[]
( as I haven't quite figured out the rule there).The only slight trickiness is we want to make sure to only constrain the deriving type's parameters to
Hashable
when it appears in the RHS (i.e. phantom type parameters should not be constrained toHashable
).I don't think it's worth using
GHC.Generics
since it's nearly as awful to use and probably much slower.The text was updated successfully, but these errors were encountered: