| 22 Oct 2025 |
 @gabyx:matrix.org | In reply to @cortsf:matrix.org
ipsavitsky234: I just made a simple program and I'd like it to be available on nixpkgs, if possible. * people can use your program also when its an output in a flake.nix, doesnt need to be in the nixpkgs repo… | 11:08:20 |
 dramforever | this seems to be a haskell thing, which i'm not really familiar about so the specifics are beyond me | 11:09:03 |
 ipsavitsky234 | yeah, you can either keep it as a flake or if you want it in nixpkgs you can take a package definition and adapt it for nixpkgs. Check out docs/language-frameworks for a good quickstart on how to package certain languages | 11:09:12 |
 cortsf | @gabyx:matrix.org: I'm also trying to provide a nixos service using this program. | 11:09:57 |
| @gabyx:matrix.org | the question is what qualifies for nixpkgs ;)? stars? usage? etc i dont. | 11:10:37 |
| cortsf | Yes, I was looking to ask about that somewhere. | 11:11:07 |
| dramforever | generally, usage | 11:12:23 |
| dramforever | we're not going to package something that's not popular among users because in nixos it is really easy to just use a package in your own nixos config | 11:13:09 |
| dramforever | including nixos service | 11:13:12 |
| dramforever | and by popular i mean, like, at least a few users right | 11:14:04 |
| dramforever | if it's just your thing that you just made, nixpkgs is not the place to advertise it | 11:14:18 |
| cortsf | Well, I thought It might become useful to others. The alternatives on nixpkgs are a bit.. outdatedd or weird. This also being haskell is very easy to customize. | 11:15:36 |
| dramforever | if you want to make it more easily available to other users, you can consider adding instructions in your readme file first | 11:15:43 |
| dramforever | so that's a problem, nixpkgs isn't really the place for software to become useful | 11:16:40 |
| dramforever | * so that's a problem, nixpkgs isn't really the place for software to become useful to others | 11:16:44 |
| cortsf | Thanks. I'll just publish a config snippet. | 11:17:44 |
| dramforever | since as mentioned, especially with flakes or something, we have the very viable alternative of users pulling your flake | 11:17:47 |
| dramforever | instead of having to go through nixpkgs | 11:17:59 |
| cortsf | It's ok, I knew this was going to be likely the case. Thanks! | 11:18:22 |
| @gabyx:matrix.org | also its good to see more nic adoption in other places ;) | 11:19:47 |
| @gabyx:matrix.org | if you go the nix flake way, once the stuff works and people find it useful you can still package to nixpkgs | 11:20:20 |
| 23 Oct 2025 |
|  @mrargoz:matrix.org joined the room. | 21:01:33 |
| 24 Oct 2025 |
| @mrargoz:matrix.org left the room. | 04:39:39 |
|  @xengi42:matrix.org left the room. | 19:34:21 |
| 25 Oct 2025 |
|  byteflavour joined the room. | 21:48:44 |
| 26 Oct 2025 |
|  @acidbong:envs.net joined the room. | 07:14:09 |
| @acidbong:envs.net | moin. i'm trying to expose my NixOS' config.i18n.glibcLocales (for the sake of being built and cached by CI) in a flake-parts way, but there's a catch
here's what I define:
{lib,self,...}: {
perSystem = {system, ...}: {
packages = {
glibcLocales-bong = lib.mkIf (
self.nixosConfigurations.bong.config.nixpkgs.hostPlatform.system == system
) self.nixosConfigurations.bong.config.i18n.glibcLocales;
};
};
}
I expect it to generate this
{
packages = {
aarch64-linux = {
...
dwm = ...;
hacksaw = ...;
...
};
x86_64-linux = {
...
dwm = ...;
glibcLocales-bong = ...;
hacksaw = ...;
...
};
};
}
but it does not: instead, when running nix flake show --all-systems, it still tries to evaluate it on a wrong system
$ nix flake show --all-systems
...
└───packages
├───aarch64-linux
│ ├───desktop-scripts: package 'desktop-scripts'
│ ├───dwm: package 'dwm-6.5'
error: The option `perSystem.aarch64-linux.packages.glibcLocales-bong' was accessed but has no value defined. Try setting the option.
before submitting to flake.parts, I wanna make sure whether they misimplement this or i misunderstood Nixpkgs modules and lib.types.attrsOf in particular | 07:16:41 |
| @acidbong:envs.net | turns out, a higher-level mkIf was the way, I did misunderstood [lazyA,a]ttrsOf | 09:05:53 |
|  Sir_Morton joined the room. | 15:33:28 |
| 27 Oct 2025 |
| @acidbong:envs.net left the room. | 09:20:03 |
