| 30 Aug 2025 |
 tsn9cp_again | Quick question about the naming convention of package/derivation versions in Nix:
If I build from a specific commit, which is present only in a specific branch, how would I name the version? AFAIK the convention for non-tagged commit on the main branch is "unstable-YYYY-MM-DD". Where would the branch fit into this? | 10:36:37 |
 K900 | Better question for #Nixpkgs / NixOS contributions | 10:37:11 |
| 31 Aug 2025 |
|  @alper-celik:matrix.org left the room. | 16:00:10 |
|  sinan changed their profile picture. | 16:25:25 |
| 2 Sep 2025 |
 Robert Hensing (roberth) | @Mic92 fyi https://github.com/NixOS/nix/issues/13898 | 14:12:20 |
| Robert Hensing (roberth) | also tagged it and the linked one for triage | 14:13:02 |
 Mic92 | Robert Hensing (roberth): nix-eval-jobs also has --apply but this is per derivation i.e. to compute additional values for the json output | 14:35:53 |
| Mic92 | For nix --apply might be a good name itself. | 14:39:32 |
 Philip Taron (UTC-8) | Opened https://github.com/NixOS/nixpkgs/pull/439645 for packaging Nix 2.28.5 (update to stable) | 17:25:00 |
| 3 Sep 2025 |
|  @ghpzin:envs.net joined the room. | 09:29:47 |
| 4 Sep 2025 |
 vcunat | FYI, a nix test regressed on master apparently about one day ago:
https://hydra.nixos.org/job/nix/master/tests.nix-copy | 09:39:44 |
| vcunat | (no idea why jobsets with flakes won't show commits) | 09:40:42 |
 Sergei Zimmerman (xokdvium) | Hm: more like it’s been flaky. Concerning | 10:03:08 |
| vcunat | No. The red circles are cancellations. | 10:24:11 |
| vcunat | Because running 600 jobs on every commit in the Nix repo would be lots of work. | 10:25:07 |
| vcunat | Ah, though I see some red crosses further back in August, too. | 10:26:12 |
