Closed andir closed 2 years ago
I have upgraded nixpkgs in #65 and fixed the Haskell errors. Can you give it a shot on both x86-64 and aarch64?
We may also need to upgrade GHC to make your build happy, though you might be able to do this yourself through overrides. Either way, please report back.
I will try and upgrade GHC to v9.0.2 after this, but v9.2 (with native aarch64 support) is off-limits until it's also available through a Stackage LTS release.
(Decided I might as well merge that, so please test against trunk.)
Looks better. Ran into some issues while compiling GHC but haven't figured out if that is just flaky due to the lack of a proper sandbox.
OK, glad to hear that it's looking better, at least. Please close the issue when you're sure.
Closing this; assuming it works well now.
It would be great if you could upgrade to a newer nixpkgs version. In particular one from the last couple of months. The reason for it being aarch64-darwin support which is becoming more and more common.
Within
npmlock2nix
we are using some for our integration tests and some users/contributors are on aarch64-darwin. Unfortunately passing a newernixpkgs
instance via thepkgs
attribute of yourdefault.nix
doesn't entirely cut it. This leads to some weird Haskell errors that I can't make any sense of (as a non-haskell developer).