I wanted to link this previous discussion about release cadence in case anyone reading this thread hasn’t seen that one: Priorities for upcoming GHC releases - #46 by Bodigrim
GitHub - haskell-CI/haskell-ci: Scripts and instructions for using CI services (e.g. Travis CI or Appveyor) with multiple GHC configurations generates the github actions files from the cabal files in the project. Eg, you add a new tested-with compiler, regenerate CI, and then that’s added to your workflows.
This conversation makes me think that we need to do a better job of catching performance regressions in fundamental libraries like text and bytestring before a GHC is released. Perhaps we could teach head.hackage to run and compare benchmark results.