14 Oct 2024 |
@jade_:matrix.org | well, i guess we get to feature detect nixpkgs harder | 20:40:58 |
@jade_:matrix.org | but this kind of sucks | 20:41:04 |
@jade_:matrix.org | technically we run user input through lowdown too but not that spicily | 20:41:20 |
puck | actually, looks like it might be okay | 20:41:21 |
puck | it only tests compilation, not execution? | 20:41:59 |
puck | it's just that the test is entirely valid for execution too | 20:42:26 |
@jade_:matrix.org | https://git.lix.systems/lix-project/lix/issues/547 btw | 20:44:13 |
Ross A. Baker | Oh, wow. Sorry for the Bruno sidetrack. I completely overlooked that it was a merge commit on nixpkgs. 🤦 | 20:45:30 |
@jade_:matrix.org | no worries, bisecting is hard | 20:47:09 |
vuimuich | Hi, I just noticed that the direnv for my flake repo does not use lix, but falls back to nix 2.18.4 can someone point me in the right direction why this happens? I did not set the nix version for direnv afaik. | 21:44:39 |
vuimuich | I have only programs.direnv.enable = true and programs.direnv.nix-direnv.enable = true configured in home-manager. | 21:49:12 |
@jade_:matrix.org | In reply to @vuimuich:matrix.org Hi, I just noticed that the direnv for my flake repo does not use lix, but falls back to nix 2.18.4 can someone point me in the right direction why this happens? I did not set the nix version for direnv afaik. this is the nix-direnv bug | 22:24:33 |
@jade_:matrix.org | https://github.com/nix-community/nix-direnv/pull/513 | 22:24:56 |
@jade_:matrix.org | https://github.com/NixOS/nixpkgs/pull/339358 | 22:25:13 |
vuimuich | what direnv 3.0.6? Somehow I am still on 2.34.0, need to check if I unintentionally pinned that or something... | 22:28:54 |
vuimuich | jade_: thanks for the links!! | 22:29:07 |
@jade_:matrix.org | nix-direnv | 22:29:15 |
@jade_:matrix.org | not direnv | 22:29:18 |
vuimuich | oh right 😊 | 22:29:42 |
@jade_:matrix.org | i am opening a 24.05 backport | 22:35:20 |
vuimuich | 🤔 I am already on nix-direnv 3.0.6 (using nixos-unstable).. but I guess I need to postpone further investigation till tomorrow. Nevertheless thanks again for your help. | 22:37:20 |
vuimuich | * 🤔 I am already on nix-direnv 3.0.6 (using nixos-unstable and it's the only version I have in my /nix/store ).. but I guess I need to postpone further investigation till tomorrow. Nevertheless thanks again for your help. | 22:39:50 |
aftix | does nix build .#foo try to find foo itself in the substituters? | 22:54:47 |
@jade_:matrix.org | i think you are asking if there is an evaluation cache that uses the internet. the answer to that is no | 22:55:24 |
@jade_:matrix.org | the only things cached are /nix/store/..., which are substituted | 22:55:39 |
aftix | I was asking if the outputs of the derivation are substituted, not evaluation | 22:56:27 |
@jade_:matrix.org | ah, yes, substitution is per derivation output | 22:57:04 |
@jade_:matrix.org | you will observe e.g. that build-only inputs never wind up on your machine if the output is available in a substituter | 22:57:32 |
@jade_:matrix.org | https://gerrit.lix.systems/c/lix/+/2068 here | 23:52:07 |
@jade_:matrix.org | * https://gerrit.lix.systems/c/lix/+/2068 here we go, that is the macOS fix | 23:52:15 |