| 26 Jun 2026 |
 hexa | my silly projection says we have like 4 more hours of work queued | 15:25:21 |
 Alyssa Ross | none from me for once — I did my musl homework early this time | 15:25:21 |
 whispers [& it/fae] | https://github.com/NixOS/nixpkgs/pull/535341 is a darwin stdenv rebuild | 15:25:34 |
 emily | yes, a few Darwin things | 15:25:38 |
| emily | should be doable by tonight | 15:25:52 |
 Grimmauld (any/all) | I am throwing a hook into llvmPackages.openmp, not sure how low in the stack that is yet | 15:25:55 |
| emily | just waiting on the build | 15:25:55 |
| emily | also Linux stdenvs will be rebuilding I think | 15:26:23 |
| emily | since we reverted GCC bump | 15:26:25 |
| emily | unless that didn't get built in the first place | 15:26:31 |
 Scrumplex | https://github.com/NixOS/nixpkgs/pull/534906
curl update should be ready for review/merge. I was able to build all tests across all curl variants as well as lix by reverting the upstream change causing the hang during checkPhase.
This will relieve us from a bunch of low to medium severity CVEs | 15:27:53 |
| hexa | nixpkgs/staging has an eval from 2h ago with only aarch64-darwin still queued | 15:28:11 |
| hexa | 478605f6e9ca | 15:28:20 |
| emily | you can cancel the Darwin on that I suppose | 15:28:37 |
| whispers [& it/fae] | i had to do no non-trivial local rebuilds for those on *-linux | 15:28:40 |
| hexa | are you sure? | 15:28:55 |
| emily | well we certainly want ^ I assume | 15:29:13 |
| hexa | ok, done | 15:29:19 |
| emily | and I assume we'll want https://github.com/NixOS/nixpkgs/pull/535508 once I confirm it builds | 15:29:31 |
| emily | (~100 builds away) | 15:29:37 |
| emily | or we could preemptively merge it and let Hydra race to see :P | 15:30:15 |
| hexa | nuh-uh | 15:30:25 |
| emily | I also have a bunch of commits to fix e.g. update scripts breaking because of the x86_64-darwin drop, but those shouldn't cause a billion rebuilds, so can merge during the cycle | 15:31:07 |
| emily | and the remaining big-rebuild cleanups I have lying around can just catch the train for the next cycle | 15:31:39 |
| emily | pretty excited to see how fast a cycle goes with new queue runner + no x86_64-darwin | 15:36:26 |
| emily | what if aarch64-darwin finishes half-way through and we need to bring x86_64-darwin back to give the builders more work? | 15:37:19 |
| emily | btw, it looks like we override supportedSystems for staging-next https://hydra.nixos.org/jobset/nixpkgs/staging-next#tabs-configuration | 15:37:59 |
| emily | that's presumably going to need to be fixed | 15:38:02 |
| hexa | removed | 15:39:51 |
| hexa | because what we have in staging as a default looks fine | 15:40:03 |
