This would, for example, allow native support in servant-auth, without the need for a separate servant-auth-rawm package or user-defined orphan instances.
Would you be open to this idea, or do you prefer maintaining servant-rawm as a separate package ?
There is some interest for shipping the
RawM
combinator as part ofservant
, see:https://github.com/haskell-servant/servant/issues/1544 https://github.com/haskell-servant/servant/pull/1551
This would, for example, allow native support in
servant-auth
, without the need for a separateservant-auth-rawm
package or user-defined orphan instances.Would you be open to this idea, or do you prefer maintaining
servant-rawm
as a separate package ?