| 15 Nov 2021 |
 Vladimír Čunát | Maybe it won't be easy to make the cycle without delaying the planned 21.11 branch-off. | 17:40:05 |
 jonringer | In reply to @vcunat:matrix.org
Maybe it won't be easy to make the cycle without delaying the planned 21.11 branch-off. Well, some of the jobs like r-updates or the glibc bump don't really need to be going on during zhf | 17:43:12 |
| Vladimír Čunát | OK, let me cancel glibc and failglob. I know those. They had very low shares set up, but let's not burden the queue scheduling now, etc. | 17:46:07 |
| Vladimír Čunát | I don't know what's with r-updates now. | 17:46:44 |
| Vladimír Čunát | trunk-combined should be mostly cached, only Hydra hasn't processed that yet. | 17:47:03 |
| jonringer | This week, and the first post-branch-off evaluations are probably the most critical | 17:47:04 |
| jonringer | Yea, I noticed that trunk-combined has dipped a lot, it will still likely need to do a lot of nixos tests, but that's fine | 17:47:36 |
 sterni | Vladimír Čunát: r-updates is due to staging-next reaching that branch | 17:48:25 |
| Vladimír Čunát | In reply to @sternenseemann:systemli.org
Vladimír Čunát: r-updates is due to staging-next reaching that branch Meaning it's also mostly cached but not "shown" yet? | 17:48:58 |
| sterni | IIRC rPackages was removed from release.nix in 2016, so staging-next causes a complete cache miss for that branch | 17:49:05 |
| sterni | oh no it has been re-enabled it seems | 17:49:29 |
| sterni | * ~~oh no it has been re-enabled it seems~~, nevermind has recurseForDerivations = false | 17:50:24 |
| sterni | * oh no it has been re-enabled it seems~~, nevermind has recurseForDerivations = false | 17:50:34 |
| Vladimír Čunát | master..r-updates isn't empty anyway. | 17:50:48 |
 hexa | In reply to @jonringer:matrix.org
When do we want to start the next staging PR? would be kind to ping before opening, so we can get some last minute things in 🙂 | 17:50:54 |
| sterni | Vladimír Čunát: yeah, but due to staging-next it is a full rebuild, otherwise I suspect some cache would still be valid… | 17:51:25 |
| sterni | but supposedly r builds are fast… https://github.com/NixOS/nixpkgs/commit/ccd1029f58a3bb9eca32d81bf3f33cb4be25cc66 | 17:51:49 |
| sterni | Vladimír Čunát: jbedo hangs out in nixos dev, in case we want to coordinate queue issues | 17:53:42 |
| jonringer | In reply to @sternenseemann:systemli.org
but supposedly r builds are fast… https://github.com/NixOS/nixpkgs/commit/ccd1029f58a3bb9eca32d81bf3f33cb4be25cc66 maybe, but 40k fast builds is still 40k builds. Same can be said about a lot of python builds | 17:57:05 |
| sterni | indeed, I don't think it came quite across as a snarky comment | 17:58:21 |
| jonringer |
so we can get some last minute things in 🙂
Sure, but this cycle is already ~3 days late
| 17:58:18 |
| jonringer | Oh, I just meant that it's still "build pressure", and it's concerning if I want to start another staging-next with ~90k builds | 17:59:12 |
| Vladimír Čunát | So... we cancel them and ping jbedo in case there's some importance to have them "now"? | 17:59:54 |
| sterni | I guess we can ask how pressing it is? | 18:00:32 |
| sterni | in any case we are still waiting on hexa's cpython update | 18:00:43 |
| jonringer | well, if we are going to delay to get last minute PRs in. I'm fine delaying a few hours. I just started my day. But I would like to go to bed with it started | 18:01:06 |
| sterni | also is there anything in here that should be included? https://github.com/NixOS/nixpkgs/pulls?q=is%3Apr+is%3Aopen+base%3Astaging | 18:01:25 |
| hexa | In reply to @jonringer:matrix.org
so we can get some last minute things in 🙂
Sure, but this cycle is already ~3 days late
this cycle won't get into 21.11 anyway, no? | 18:08:00 |
| jonringer | incorrect | 18:08:27 |
| hexa | oops | 18:08:37 |
