| 14 Jan 2025 |
 alexfmpe | The only decision left is whether to call it nax or tix | 13:33:43 |
|  Luke joined the room. | 20:47:48 |
| 16 Jan 2025 |
 sterni (he/him) | https://github.com/NixOS/nixpkgs/pull/374323 | 15:09:09 |
| sterni (he/him) | I'd want to merge that soon and trigger another eval, so we can generate a hydra report with less noise. | 15:15:29 |
| alexfmpe | did not expect HLS to depend on clay of all things | 23:06:45 |
| alexfmpe | failed eval due to some "time to remove me" override on clay | 23:07:02 |
| alexfmpe | let's see what else it fails on | 23:07:11 |
| alexfmpe | via stan | 23:07:20 |
 maralorn | In reply to @alexfmpe:matrix.org
did not expect HLS to depend on clay of all things Yeah, @manuelbaerenz:matrix.org stumbled over that recently when iterating on the clay repo flake. 😂 | 23:09:15 |
| 17 Jan 2025 |
| alexfmpe | "why am I building 9.6" he says, not noticing supportedGhcVersions ? [ "96" ] | 00:37:17 |
| alexfmpe | if we're defaulting to singleton-of-blessed-version, can we get this in sync somehow? | 00:37:50 |
| alexfmpe | e.g.
nix-repl> lib.strings.replaceStrings ["."] [""] (lib.versions.majorMinor haskellPackages.ghc.version)
"98"
| 00:43:04 |
| alexfmpe | I'm wary of fixed point demons | 00:43:35 |
| alexfmpe | but it's not obviously wrong | 00:43:40 |
 Manuel Bärenz | In reply to @maralorn:maralorn.de
Yeah, @manuelbaerenz:matrix.org stumbled over that recently when iterating on the clay repo flake. 😂 Since then I feel vindicated for co-maintaining it | 06:47:48 |
| sterni (he/him) | https://mail.haskell.org/pipermail/glasgow-haskell-users/2025-January/027255.html | 13:12:56 |
 thirdofmay18081814goya | wasn't there a nix embedding in haskell? | 15:02:27 |
| thirdofmay18081814goya | today is the day I've decided, I will not write one more line of nix, unless it is in the process of creating a haskell-to-nix compiler | 15:02:56 |
| maralorn | Not really, there is a purescript to nix compiler | 15:09:48 |
| maralorn | And there is hnix which is a nix implementation in Haskell. | 15:10:07 |
| thirdofmay18081814goya | i just want the simplest of typing systems, god please lord give me types I will die | 15:10:31 |
| maralorn | Which can probably be used as a library to do nix stuff in Haskell. | 15:10:32 |
| maralorn | You might wanna have a look how far nickel-lang.org is usable as a nix drop-in. | 15:11:11 |
| thirdofmay18081814goya | oh two very good suggestions | 15:11:24 |
| thirdofmay18081814goya | thank you! | 15:11:25 |
| maralorn | But while I totally share your pain I am quite sceptical that any replacement is viable for most use cases without significant effort at this point. | 15:12:24 |
| thirdofmay18081814goya | modeling an attempt to do this on how typescript did it might give a shimmer of hope | 15:13:30 |
| thirdofmay18081814goya | or at the very least I want to delegate any processing of lists and attribute sets to another language | 15:14:22 |
| thirdofmay18081814goya | but one thing that is absolutely unacceptable to me is the ... input | 15:16:55 |
| thirdofmay18081814goya | which wouldn't be necessary if functions could share type information at their call location | 15:17:39 |
