| 27 Apr 2025 |
 alexfmpe | I think the force bit in force push is lazy | 17:58:57 |
|  @denizalpd:matrix.org left the room. | 18:00:29 |
| 28 Apr 2025 |
|  chvp left the room. | 14:30:17 |
| 29 Apr 2025 |
|  ortolanbunting3002 joined the room. | 01:26:35 |
 woobilicious | man I'm having this super weird issue where typed-process in ghci via nix-shell has the Process type, but nix shell for my haskell script doesn't | 06:15:28 |
 Alex | In reply to @woobilicious:matrix.org
man I'm having this super weird issue where typed-process in ghci via nix-shell has the Process type, but nix shell for my haskell script doesn't Different Nixpkgs versions? | 07:03:34 |
| woobilicious | should be the same version, I think it's just a quirk of the package, I was using it to pattern match so I just changed my code, which was required for the next step anyway. | 07:08:52 |
| woobilicious | Converting a bash script to Haskell just because I can lol. | 07:10:36 |
|  @malteneuss:matrix.org joined the room. | 08:48:25 |
| @malteneuss:matrix.org left the room. | 10:33:14 |
| 30 Apr 2025 |
|  @devalot:matrix.org left the room. | 09:06:01 |
 JoelMcCracken | curious, on hackage it appears that typed-process had a Process type since the first published version | 13:32:52 |
| @malteneuss:matrix.org joined the room. | 16:56:29 |
| 1 May 2025 |
 thirdofmay18081814goya | Is there a way to package a cabal-install such that it comes with a user config? | 16:59:58 |
| thirdofmay18081814goya | i.e. with what would be usually in $HOME/.cabal/config | 17:00:29 |
| thirdofmay18081814goya | probably the way to do this is to have nix-direnv read a shell with a shellHook specifying a path to a config right | 17:09:52 |
 sterni (he/him) | runCommand "my-cabal" { nativeBuildInputs=[makeWrapper];} "mkdir -p $out/bin; makeWrapper ${cabal-install}/bin/cabal "$out/bin/cabal" --add-flags "--config-file ${my-config}" probably | 20:06:28 |
|  rosariopulella changed their display name from Rosario Pulella to Rosuavio. | 20:08:02 |
 maralorn | Oh, boy … I am really glad that we have an autoformater now. It is just sad that it feels like it is optimizing for maximal line count. which just means that my eye can see less code at a time and needs more caching and time to read and understand a function. https://github.com/NixOS/nixpkgs/pull/403343/files#diff-06cdf278a2cf7f0e7793a4a8d19cbdd6e6a3a3722482708833ef9270b65d7d56R2990-R3006 | 22:07:47 |
| maralorn | Also I assume that 99% of contributors don’t know that the formatter preserves some newlines which actually stay away when you remove them. | 22:09:06 |
| alexfmpe | Yeah I don't much care for the specific format conventions chosen | 22:19:23 |
| alexfmpe | hackage-packages.nix is almost a million lines now | 22:19:59 |
| alexfmpe | Need to page down a few times even after search-navigation to "foopackage" | 22:20:31 |
| maralorn | I lobbied a lot for a more compact layout, but I got overruled (which is fair). | 22:33:37 |
| maralorn | Especially I understand that diffs are a large concern in nixpkgs. | 22:34:31 |
| alexfmpe | Ah well, generated files get hidden diffs by default | 22:36:00 |
| alexfmpe | Though in this case you'd get smaller diffs? | 22:36:17 |
| alexfmpe | With the old cabal2nix output, having a few new deps would shift all the lines no? | 22:36:38 |
| alexfmpe | * With the old cabal2nix output, having a few new deps would shift all the lines after them no? | 22:36:48 |
| alexfmpe | Speaking of hackage-packages.nix, why are there so many doHaddock = false in there? | 22:37:38 |
