Hello, QA also finds quite a few new build failures on x86_64: https://qa.guix.gnu.org/branch/haskell-team/package-changes?x86_64-linux-change=broken&x86_64-linux-change=still-failing&x86_64-linux-change=unknown-to-failing&x86_64-linux-change=new-failing Unfortunately lots of ghc-* packages.
In case this page disappears as other branches become active, here is one example: https://data.qa.guix.gnu.org/gnu/store/i48030qa0y82mrlklry12kn9akmpnglf-ghc-highlighting-kate-0.6.4.drv https://bordeaux.guix.gnu.org/build/18649055-f72c-415c-8078-2a39842844c7/log Text/Highlighting/Kate/Common.hs:233:27: error: [GHC-88464] Variable not in scope: unless :: Bool -> Text.Parsec.Prim.ParsecT s0 SyntaxState m0 () -> Text.Parsec.Prim.ParsecT [Char] SyntaxState ghc-internal-9.1002.0:GHC.Internal.Data.Functor.Identity.Identity a0 | 233 | unless (null xs) $ | ^^^^^^ I wonder whether we should for the time being limit the scope of the branch to the addition of new GHC versions, getting them all to build also on aarch64, and reserve the bump to a new default version for the next iteration? I suppose that the different "fix build" commits would still be compatible with this approach. By the way, the first line of the commit messages should end with a period ('.'). Andreas
