| 29 Jul 2024 |
phaer | Ah, that makes sense! Guess it's worth one extra API call to figure out the project id by name is worth it anyway ;) | 13:48:27 |
phaer | * Ah, that makes sense! Guess it's worth one extra API call to figure out the project id by name anyway ;) | 13:48:38 |
Jonas Chevalier | Yeah, the UI doesn't make it obvious that when you change the setting, it actually pushes a git commit in the gh-branches page with that CNAME file. | 16:17:11 |
| 30 Jul 2024 |
| varvir joined the room. | 08:31:41 |
varvir | Hi! I want to get some help about nix-direnv. I followed the installation guide at flake support. But I encountered the error error: path '/nix/store/0fzhpkbnbbc0ywb2hp4l8bxfpqxr2fjc-source/flakedirenv/flake.nix' does not exist. | 08:33:22 |
varvir | Only using nix method works fine. | 08:34:09 |
K900 | Add flake.nix to git | 08:35:26 |
varvir | oh Thanks! it works. | 08:37:39 |
| 1 Aug 2024 |
| @thescientiac:matrix.org left the room. | 05:16:06 |
| Vika (she/her) changed their profile picture. | 11:50:13 |
| 2 Aug 2024 |
| @travis-staton:matrix.org joined the room. | 15:45:53 |
Matt Sturgeon | Just tried using nix-direnv for the first time; it seems it doesn't work with git worktrees?
$ cd nixvim/docs_raw_submodule/
direnv: error /home/matt/nixvim/docs_raw_submodule/.envrc is blocked. Run `direnv allow` to approve its content
$ direnv allow
direnv: loading ~/nixvim/docs_raw_submodule/.envrc
direnv: using flake
error:
… while fetching the input 'git+file:///home/matt/nixvim/docs_raw_submodule'
error: '/home/matt/nixvim/docs_raw_submodule' is not a Git repository
direnv: nix-direnv: Evaluating current devShell failed. Falling back to previous environment!
direnv: export +NIX_DIRENV_DID_FALLBACK ~PATH
In this example ~/nixvim is a "bare" git repo and docs_raw_submodule is a worktree created using git worktree add
| 20:38:22 |
| 3 Aug 2024 |
Matt Sturgeon | Seems to be working now after a reboot | 15:09:55 |
toonn | StackOverflow recommends against using bare repos like that, FYI. I checkout an empty commit instead. | 15:31:58 |
Matt Sturgeon | I used to have the default branch checked out at the top-level and then have feature branches checked out in nested worktrees, but I found it annoying how the nested worktrees showed up as untracked files in the top-level worktree, figured a bare repo was cleaner.
Do you recall why someone on stackoverflow recommended against using a bare checkout? Was it just because some tooling might be expecting a .git dir? I find most tools use libgit or something similar, so it usually isn't an issue.
| 15:44:36 |
toonn | MattSturgeon: This answer has the reasoning at the end, https://stackoverflow.com/a/54408181 | 15:56:44 |
toonn | I don't understand the refspec details so I figured I'd better not mess with it. | 15:57:07 |
toonn | So I have a git alias to checkout an empty commit. Gets pretty much the same result without any caveats I don't properly understand. | 15:57:36 |
| 6 Aug 2024 |
| @cole-h:matrix.org left the room. | 22:16:29 |
| 7 Aug 2024 |
| @taffisher:matrix.org left the room. | 19:13:01 |
| 8 Aug 2024 |
Matt Sturgeon | IDK if the sheer quantity of builds we're currently running in nixvim is contributing, but we're running into buildbot issues with cross-comp/available builders rather frequently, leading to us having to rebuild to get a green checkmark (putting even more strain on resources)...
e.g. this is from a recent build (from this eval):
Failed to find a machine for remote build!
derivation: ycm30n2fwjwycmx3hi21gvn1q4kr04yy-plugins-ui-zen-mode.drv
required (system, features): (aarch64-linux, [])
2 available machines:
(systems, maxjobs, supportedFeatures, mandatoryFeatures)
([aarch64-linux], 80, [benchmark, big-parallel, gccarch-armv8-a, kvm, nixos-test], [])
([aarch64-darwin, x86_64-darwin], 8, [big-parallel], [])
error: a 'aarch64-linux' with features {} is required to build '/nix/store/ycm30n2fwjwycmx3hi21gvn1q4kr04yy-plugins-ui-zen-mode.drv', but I am a 'x86_64-linux' with features {benchmark, big-parallel, kvm, nixos-test}
| 21:16:48 |
| Austin Horstman joined the room. | 21:18:11 |
| 9 Aug 2024 |
zowoq | I've seen that occur a few times in other projects on buildbot as well. For now the rebuilds aren't too much strain on our resources. There are plans to hopefully make this less of an issue, e.g. smarter scheduling and making no op rebuilds faster. | 01:43:52 |