| 21 Jan 2025 |
 emily | something up with the AArch64 box?
yuyuko:~
❭ ssh aarch64-build-box.nix-community.org
Connection closed by 65.109.82.88 port 22
| 19:23:27 |
|  @rootname:matrix.org removed their display name rootname. | 19:55:13 |
| @rootname:matrix.org left the room. | 19:59:54 |
| emily | Gaétan Lepage: load average on the x86 builder is >200 right now | 20:17:36 |
 Gaétan Lepage | I got disconnected a few hourse ago | 20:17:41 |
| Gaétan Lepage | I don't think it's my fault this time | 20:17:49 |
| Gaétan Lepage | Have you managed to connect ? | 20:19:27 |
| emily | not to the AArch64 builder | 20:21:07 |
 zowoq | It's back. | 21:21:47 |
| Gaétan Lepage | Thanks ! | 21:22:35 |
 hexa | how many more times do we need to do this dance? | 21:41:51 |
| hexa | (apparently my firefox build was killed) | 21:43:23 |
 Roberto Abdelkader Martínez Pérez | Hey nix-community folks! 👋
I'm the maintainer of nix-community/autofirma-nix. Since moving the project under the nix-community organization, I'm planning to migrate our users from our old Cachix cache to the one provided by the organization.
I'm curious: has anyone here gone through a similar migration of users for a project? What strategies or best practices did you find helpful during that transition? Any lessons learned or suggestions on how to smoothly guide users through such a migration would be greatly appreciated!
Thanks so much for sharing your experiences and advice! 😊
| 23:23:52 |
| 22 Jan 2025 |
| emily | Mic92: I see you cut the last few releases of nix-init, would you be able to take a look at getting one for https://github.com/nix-community/nix-init/pull/589? the upcoming libgit2 bump in Nixpkgs is about to break nix-init otherwise | 23:12:18 |
 Mic92 | emily: I have the feeling we have a regression in nixpkgs in general because we are missing .pc files in many cases: https://github.com/nix-community/nix-init/actions/runs/12918789306/job/36028026452?pr=589#step:5:956 | 23:17:20 |
| Mic92 | I have seen this in nix as well | 23:17:29 |
| emily | no, it's just broken because we haven't bumped libgit2 in Nixpkgs yet | 23:20:48 |
| emily | that's why I had to test with --override-input | 23:20:53 |
| emily | it's a catch-22 since by the time it hits the channels nix-init will be broken | 23:21:03 |
| emily | so ideally we'd get a nix-init release cut before the staging-next (as the main consumer of releases is presumably Nixpkgs) | 23:21:29 |
| emily | I could pin the flake input to the PR I suppose 😅 | 23:22:25 |
| Mic92 | emily: you mean staging-next? | 23:23:01 |
| Mic92 | I think I would prefer pinning that over breaking ci | 23:23:13 |
| emily | it's not in -next yet (it's not even in staging) | 23:23:32 |
| emily | we could merge into staging and then pin staging, but then CI would be building stdenv | 23:23:41 |
| emily | pinning to the PR head produces a working build though | 23:23:48 |
| Mic92 | Ok. I guess it would be fine to bump a nix-init release once we it hits staging-next. | 23:25:44 |
| Mic92 | No one will be mad if it's broken in staging for a while. | 23:26:51 |
| emily | fair enough | 23:28:24 |
| emily | pinning staging-next in the flake feels a bit weird though | 23:28:40 |
