| 7 Oct 2025 |
lovesegfault | okay, 2.32 is branched off, version has been bumped, can someone hit merge on https://github.com/NixOS/nix/pull/14147 for me now? | 21:07:29 |
lovesegfault | * okay, 2.32 is branched off, version has been bumped, can someone hit merge on https://github.com/NixOS/nix/pull/14147 for me? | 21:07:34 |
Sergei Zimmerman (xokdvium) | I really wish we could clean github actions before doing this | 21:07:59 |
Sergei Zimmerman (xokdvium) | Now they are a total mess | 21:08:08 |
Sergei Zimmerman (xokdvium) | Like vm tests should just be part of regular tests | 21:08:38 |
lovesegfault | my hope is that change is just reverted after we're done churning this code | 21:08:45 |
Sergei Zimmerman (xokdvium) | But also you are not running s3 tests in ci anyway? | 21:08:55 |
Sergei Zimmerman (xokdvium) | I'd hope that it's not. We just should run vm/s3 tests all the time | 21:09:29 |
lovesegfault | anyway, the main benefit is that in while we work on s3 you don't have to trust me that i tested that it still builds with s3 disabled or with the legacy s3 enabled :P | 21:10:51 |
lovesegfault | the variant testing we can revert | 21:10:59 |
Sergei Zimmerman (xokdvium) | I'm very much in favor of running tests/builds for s3 store. We just shouldn't have 2 new matrices for this and even more copy-pasta | 21:13:41 |
Sergei Zimmerman (xokdvium) | If we could just add the vm tests to regular tests and add a new config to build with the new s3 that'd be great | 21:14:15 |
Sergei Zimmerman (xokdvium) | Then we'd get the benefit of running vm tests with sanitizers as well | 21:14:49 |
lovesegfault | 🫠| 21:14:53 |
lovesegfault | i won't have time to do that for a while, and i'm spending too much time on this curl work as it is | 21:15:39 |
Sergei Zimmerman (xokdvium) | Sure, I had a draft branch for the cleanup. I wasn't suggesting you needed to tackle that | 21:17:51 |
lovesegfault | can you point me to that? | 21:18:32 |
Sergei Zimmerman (xokdvium) | Don't think I've pushed it anywhere yet | 21:19:01 |
Philip Taron (UTC-8) | I'll get a PR spun up for Nixpkgs | 21:24:38 |
Sergei Zimmerman (xokdvium) | Thanks! There's some flags you'll need to yeet to not tank performance | 21:27:06 |
Sergei Zimmerman (xokdvium) | https://github.com/NixOS/nix/pull/14105/commits/c1f805b8569d1f66aed813a3b49820936618c9d5 | 21:27:24 |
Philip Taron (UTC-8) | Is that suitable for all Nix versions, or just 2.32? | 21:28:04 |
Philip Taron (UTC-8) | It's a bit hard to apply that solely to Nix 2.32, so I think it'll get applied to even stable. | 21:32:13 |
Sergei Zimmerman (xokdvium) | I'm not sure it's safe to apply to other nix versions tbh. I've run into unfortunate UB that this surfaced. I'd gate this for 2.32 and upwards | 21:34:16 |
Sergei Zimmerman (xokdvium) | Mic92: Now that we have the backport action what's the next steps for github merge queues? Org issue? | 21:43:37 |
Philip Taron (UTC-8) | https://github.com/NixOS/nixpkgs/pull/449668 | 21:52:34 |
lovesegfault | Sergei Zimmerman (xokdvium): can you elaborate on the ci improvements you'd like to see? | 23:31:02 |
Sergei Zimmerman (xokdvium) | Well, first it'd be nice if vm_tests were just part of tests. I've already done the refactor to make the nixos tests accept nixComponents. There we can enable s3-binary-cache-store test + evalNixpkgs | 23:32:20 |
Sergei Zimmerman (xokdvium) | cc getchoo | 23:32:35 |
lovesegfault | okay, let me see what i can do | 23:35:28 |