| 12 Apr 2024 |
 maralorn | And tabcompletion is worse. | 08:02:06 |
| maralorn | So making it the standard would be problematic. | 08:02:31 |
 @shalokshalom:kde.org | i see | 08:03:32 |
| maralorn | * Replacing the nix command by nom is an unsolved problem. | 08:04:07 |
| maralorn | Besides that recommend away it's nearly always the better choice IMO. | 08:05:00 |
| @shalokshalom:kde.org | Considering those are the only two reasons, I think 'not being able to support all commands' can be solved by simply doing a fallback | 08:14:41 |
| @shalokshalom:kde.org | and tabcompletion would remain as the only holdback? | 08:14:59 |
| @shalokshalom:kde.org | Probably showing my ignorance here, but are the auto completions even different from nix? wouldnt you be able to inherit them from nix? | 08:17:42 |
| maralorn | In reply to @shalokshalom:kde.org
Considering those are the only two reasons, I think 'not being able to support all commands' can be solved by simply doing a fallback Yes, there is an issue about that. | 08:18:22 |
| maralorn | In reply to @shalokshalom:kde.org
Probably showing my ignorance here, but are the auto completions even different from nix? wouldnt you be able to inherit them from nix? Basically yes. I just don't know how to. | 08:19:12 |
| @shalokshalom:kde.org | I do this with my shell, but I guess this is no option here :p | 08:21:50 |
| @shalokshalom:kde.org | Also: Congratulations for breaking the mythological spell, and producing something useful with Haskell :D | 08:23:14 |
| maralorn | There are 2 problems here: 1. Making the nom binary behave like nix. 2. Replacing nix by nom. | 08:24:17 |
| maralorn | And the second step is dangerous because nom discovers nix from path and it shouldn’t call itself recursively. | 08:24:52 |
| maralorn | But a shell alias should be fine. | 08:25:05 |
| @shalokshalom:kde.org | in fish, you can simply do --wraps "nix" https://codeberg.org/ShalokShalom/add-and-friends/src/branch/main/config.fish#L38 | 08:29:40 |
| @shalokshalom:kde.org | But I dont think bash offers this | 08:30:03 |
| maralorn | In reply to @shalokshalom:kde.org
in fish, you can simply do --wraps "nix" https://codeberg.org/ShalokShalom/add-and-friends/src/branch/main/config.fish#L38 Cool | 08:57:04 |
| maralorn | In reply to @shalokshalom:kde.org
Also: Congratulations for breaking the mythological spell, and producing something useful with Haskell :D I am certainly not the first. | 08:57:29 |
 bandithedoge | In reply to @maralorn:maralorn.de
I am certainly not the first. can't really think of anything other than pandoc and xmonad lol | 12:08:51 |
| maralorn | git-annex, hledger, shellcheck, elm, purescript, agda, darcs, nixfmt, niv, cachix, hercules-ci, … | 13:04:01 |
| maralorn | > rg haskell pkgs/top-level/all-packages.nix | wc -l
122
| 13:04:26 |
| maralorn | That is an overapproximation. | 13:05:10 |
| maralorn | Let’s say 100 Haskell apss. | 13:05:21 |
| maralorn | Maybe a third of them are in some way Haskell tooling. | 13:06:04 |
| maralorn | So maybe 70 non-navel-gazing Haskell programs. | 13:07:12 |
| maralorn | And they are not all toy-language compilers either. 😄 | 13:07:59 |
| @shalokshalom:kde.org | guess the ghc could be named too ;) | 13:32:03 |
| @shalokshalom:kde.org | I think Elm is one of my favorite Haskell programs | 13:32:28 |
| @shalokshalom:kde.org | since its not written in the typical Haskell maniac style ^^ | 13:32:43 |
