| 18 Jun 2026 |
 hexa | Redacted or Malformed Event | 01:08:30 |
 debtquity | will just invoking the problematic downstream packages work?
https://github.com/NixOS/nixpkgs/blob/d1a55f0c7b921d541c6142994c96fbd7f1fa562e/pkgs/by-name/va/valkey/package.nix#L110-L113
passthru = {
tests = {
redis = nixosTests.redis;
python3-valkey = callPackage python3.pkgs.valkey;
};
serverBin = "valkey-server";
};
???
| 01:26:27 |
| debtquity | surprised there's not much there at the moment, guess the previous maintainers had it in their head? | 01:28:00 |
 Ryan Burns | tried a nixpkgs-review and it OOMed after about 125 builds, but didn't see any failures other than the preexisting python3.pkgs.jupytext from that sample. Will try again later | 01:34:42 |
 guillaume | I can try a local nixpkgs review if it's needed | 06:56:31 |
| guillaume | I have a lot of ram ahhaa | 06:56:45 |
| Ryan Burns | I'm having more luck trying it on a less anemic machine than my laptop, haha :)
More eyes are always appreciated so feel free to beat me to the punch if you've got spare cycles but I'll be posting a proper review report soon. | 06:59:03 |
| Ryan Burns | nixpkgs-review is done and seems OK:
https://github.com/NixOS/nixpkgs/pull/532836#issuecomment-4739270789
⏩ 10 packages marked as broken and skipped:
❌ 3 packages failed to build:
python313Packages.pyinterp
python314Packages.opensfm
python314Packages.pyinterp
✅ 478 packages built:
| 07:50:29 |
 Grimmauld (any/all) | you really don't want to nixpkgs-review staging stuff, that is a glorious waste of resources | 08:08:36 |
| Grimmauld (any/all) | you get the exact rebuild count from the actions report | 08:09:01 |
| Grimmauld (any/all) | just build a few packages you know have been problematic in the past | 08:09:19 |
| Ryan Burns | We were planning to send it to master + backport, so I'm not picking up any unrelated staging builds in this report | 08:10:20 |
| Ryan Burns | I can definitely send it to staging like a typical stabilization PR, just thought it might be cleaner to do it this way for an ordinary package bump that master'll want too | 08:11:58 |
| Grimmauld (any/all) | even if you're sending it to master, 500 is quite expensive... | 08:19:13 |
| Grimmauld (any/all) | like, you don't want to nixpkgs-review kind of expensive, not necessarily staging kind of expensive | 08:19:27 |
| Ryan Burns | Point taken, just wanted to be extra sure here and had the cycles to spare, I wouldn't typically set the bar that high for someone else's PR. Just got a little overly optimistic trying to run such a nixpkgs-review on my laptop 😁 | 08:23:06 |
 Sapii/Saperson | there is https://github.com/Defelo/nixpkgs-review-gha which uses github's free runners and also does darwin | 16:17:45 |
|  whispers [& it/fae] changed their display name from whispers [& it/fae] to meow meow. | 18:46:19 |
| whispers [& it/fae] changed their display name from meow meow to whispers [& it/fae]. | 19:11:57 |
|  tea☆ joined the room. | 21:00:47 |
| tea☆ | I got a quick question regarding this PR: https://github.com/NixOS/nixpkgs/pull/514727
This has been causing build failures for wf-config (among others) for a quite a while now. I get that it can take some time for it to be merged into staging-next, but is there any way to speed this up next time so people can build their system?
(like creating a PR to skip doctests completely and then enable them again once the fix landed in unstable?) | 21:04:59 |
| hexa | hi, that would be now | 22:37:26 |
| 19 Jun 2026 |
 Vladimír Čunát | I'd hope to get this to nixos-26.05 channel by the following Monday. And hopefully a couple days after that into nixos-unstable. | 03:31:37 |
 leona | huh what happened with staging-next-26.05? Why does hydra show ~136k builds now instead of <50k | 05:23:28 |
| Vladimír Čunát | They're not real rebuilds. | 05:24:58 |
| Vladimír Čunát | It's just a consequence of migrating to the new queue-runner. | 05:25:20 |
| Vladimír Čunát | We've seen it in various jobsets yesterday. | 05:25:50 |
| Vladimír Čunát | * We've seen it in various jobsets during yesterday. | 05:25:57 |
| Vladimír Čunát | leona: in particular note that the rebuilds go back to June 06. | 05:28:08 |
| Vladimír Čunát | The errors like step 1 in https://hydra.nixos.org/build/331623748#tabs-buildsteps are really unhelpful. | 06:11:41 |
