Re: [Haskell-cafe] ghc-9 needs type declaration where ghc-8 didn't (with RankNTypes)

The 9.0 upgrade may help: https://gitlab.haskell.org/ghc/ghc/-/wikis/migration/9.0

Thanks for that pointer! So, what I observed is an effect of https://gitlab.haskell.org/ghc/ghc/-/wikis/migration/9.0#simplified-subsumpt... ? I was browsing that, but could not make the connection. "nested foralls" are mentioned only together with GADT/instance/deriving, none of which is in my example. - J.