| 3 Nov 2025 |
 K900 | ^ | 12:20:31 |
| K900 | I think the correct resolution is to move the newly added version check down to meta.broken as Jo did in their big commit | 12:21:02 |
 Grimmauld (any/all) | Can't we just cherry-pick https://github.com/NixOS/nixpkgs/commit/89fd9025cfae7cab49ea36134640952ad553c4fa to staging-next individually and then accept the staging-next version explicitly for the conflicting files? | 12:21:04 |
| K900 | And the rest already applies cleanly | 12:21:47 |
| K900 | https://github.com/NixOS/nixpkgs/commit/89fd9025cfae7cab49ea36134640952ad553c4fa#diff-87747fec8c7dfe5d924b47da9425f32fdc332203161b00ff7f754389d86b66be is the only change that doesn't | 12:22:06 |
 StepBroBD | problem solved? | 12:22:23 |
| StepBroBD | scared me for a bit | 12:22:30 |
| K900 | Well the problem will be solved if you agree this is the correct thing to do | 12:22:49 |
| K900 | And then I can just commit and push it | 12:22:54 |
| StepBroBD | yes | 12:23:01 |
| K900 | OK cool | 12:23:05 |
| K900 | Thanks | 12:23:05 |
| StepBroBD | thank you for your service | 12:23:16 |
| StepBroBD | hope i dont mess up again | 12:23:22 |
| K900 | Again you did not mess up | 12:23:30 |
| K900 | This is normal staging process | 12:23:35 |
| StepBroBD | first time joining staging workflow
i'll keep lurk around here 👀 | 12:24:20 |
| K900 | @Wolfgang Walther had some ideas in mind to reorganize the process so that there's no manual conflict resolution involved on our part | 12:24:25 |
| K900 | And instead you have to submit another PR to staging branches with the conflicts resolved correctly | 12:24:42 |
|  mtvkxsw9tngk joined the room. | 12:24:43 |
 Wolfgang Walther | Emily had those ideas and I just kept nodding. | 12:25:38 |
| K900 | But nothing implemented yet and honestly I am not super convinced it's the way to go because it's going to really confuse folks who are not familiar with staging and I'm getting carried away | 12:25:43 |
| K900 | Anyway all merged now and the automated trains can resume running cleanly, thanks for the help | 12:26:15 |
 ivy | how long does a staging cycle normally take? | 12:27:05 |
| K900 | Depends on how broken things are | 12:27:29 |
| K900 | And how much work ends up being scrapped due to some underlying changes being broken | 12:27:44 |
| K900 | Usually a week or so for a full world rebuild where nothing goes horribly wrong | 12:28:00 |
| K900 | Maybe a bit more these days since we build more things and also don't have quite as much capacity | 12:28:38 |
| Grimmauld (any/all) | did you say coreutils rebuild because less is broken? | 12:34:24 |
| K900 | I think the worst one we've had was a GCC rebuild like a week into a cycle? | 12:34:51 |
