| 5 Jun 2021 |
 sterni (he/him) | there's your motivation :) | 14:24:41 |
 maralorn | I now cap, I doubt they will ever maintain a Haskell package.^^ | 14:25:15 |
| sterni (he/him) | In reply to @maralorn:maralorn.de
Have you considered replacing [ lib.maintainers.a lib.maintainers.b ] with builtins.attrValues { inherit (lib.maintainers) a b; }? more hassle to implement plus extra evaluation cause which is not negliable I think | 14:25:16 |
| maralorn | In reply to @sternenseemann:systemli.org
more hassle to implement plus extra evaluation cause which is not negliable I think We can increase the hassle by only doing it, when it’s more than one maintainer. | 14:26:02 |
| maralorn | I think that way it would be pretty cheap too. | 14:26:17 |
| sterni (he/him) | I'm not up for touching Language.Nix.PrettyPrinting again though :p | 14:28:26 |
| sterni (he/him) | it's a really oddly designed API and although it's supposedly internal it is used in cabal2nix everywhere so I wasn't able to break it :( | 14:28:50 |
| sterni (he/him) | my dream scenario currently btw would be cabal2nix, distribution-nixpkgs and language-nix in one cabal workspace plus jailbreak-cabal transferred to the NixOS org | 14:30:24 |
| sterni (he/him) | also the ability to release new versions :D | 14:30:34 |
| maralorn | Open an issue asking for transfer and the hackage rights. | 14:33:49 |
| sterni (he/him) | yeah putting that on my list :) | 14:34:06 |
 cdepillabout | In reply to @sternenseemann:systemli.org
in other news: https://github.com/peti/distribution-nixpkgs/pull/11 You might want to add the motivation for this change in a comment so that peti understands it is a good change. | 14:47:54 |
| sterni (he/him) | Uh yeah | 14:48:11 |
| sterni (he/him) | gonna do that later and I'm gonna fix the CI finally | 14:48:20 |
| sterni (he/him) | I think I don't want to get the nix-instantiate call working inside haskell-ci that is much too much hassle | 14:48:39 |
| cdepillabout | Or just get peti to transfer the repo to NixOS. I'd be surprised if he wanted to continue maintaining that library if he has basically stepped down from NixOS stuff. | 14:48:42 |
| sterni (he/him) | yeah I think it is just a matter of nudging | 14:48:58 |
| sterni (he/him) | I think transferring is the right call since it's the least work for him | 14:49:20 |
| sterni (he/him) | having all the cabal2nix related stuff in a monorepo-ish thing would be nice | 14:49:31 |
| sterni (he/him) | but you can have a local workspace group as well probably | 14:49:40 |
| cdepillabout | Ah, yeah, having that stuff in a monorepo would be nice | 14:50:06 |
| cdepillabout | In reply to @cdepillabout:matrix.org
You might want to add the motivation for this change in a comment so that peti understands it is a good change. Oh, I guess just linking to https://github.com/NixOS/cabal2nix/issues/499 would probably work. | 14:52:50 |
| maralorn | sterni (he/him): What did you change with your force push? https://github.com/NixOS/nixpkgs/pull/125798 the commit message still says "haskell-packages" instead of "hackage-packages" | 14:59:32 |
| sterni (he/him) | ohh that's what you meant | 14:59:48 |
| sterni (he/him) | quoted the $HACKAGE2NIX var mostly in case someone has it in a path with spaces | 15:00:54 |
| maralorn | Good idea. | 15:02:58 |
 Las | Since GHC 9 won't be switched to as a default, would just compiling the GHC 9 packages in Hydra be fine? | 16:58:00 |
| maralorn | Las: Historically we have done that for a few select packages. We currently have jobs for cabal2nix, hoogle and maybe a few others. | 16:59:16 |
| maralorn | But maye outside a short transition period I don‘t think enabling it for all packages would be an option. | 16:59:52 |
| maralorn | * But (maybe outside a very short transition period) I don‘t think enabling it for all packages would be an option. | 17:00:19 |
