Release Management | 332 Members | |
| 25.11 "Xantusia" | https://nixos.github.io/release-wiki/Home.html | 91 Servers |
| Sender | Message | Time |
|---|---|---|
| 24 Apr 2025 | ||
| ah, https://github.com/NixOS/nixpkgs/pull/401428 has a patch. | 12:43:50 | |
| 25 Apr 2025 | ||
| leona: https://github.com/NixOS/nixpkgs/pull/397932 is technically a breaking change but I'd like to get it into 25.05 since it means next system update I do, I could try Steam on Apple Silicon without doing it so hacky. Reason why I think it's fine, it's just a rename and it bumps to a newer version that can actually work. | 02:25:53 | |
In reply to @rosscomputerguy:matrix.orgOkay for me | 08:28:01 | |
| 29 Apr 2025 | ||
| For some reason the unstable channel was published as Apprently from . Was this some bug/mishap in the release process? It’s breaking some scripts I have locally :D My understanding is that the unstable channel should always be pre-release. Only the beta branch for a stable release should be published as beta. Any idea what happened here? | 09:43:15 | |
| This was a manual mistake in the release process. There is some conversation about this a month ago in this channel | 09:44:03 | |
i.e. the change to beta was committed to master and release-24.11 instead of only to release-24.11 | 09:45:03 | |
| Cool. makes sense | 09:46:07 | |
| 14:41:28 | ||
| Does anyone have an opinion on whether to start tomorrow with ZHF before the last breaking changes from staging-next are merged or wait for that until staging-next is merged? It feels hard to decide unfortunately :/ | 16:27:07 | |
| I don't think it really makes sense to start ZHF before it's really known what's "really" broken. But we might to have some more days | 16:29:00 | |
They might start fixing failures on staging-next 😆 | 16:47:21 | |
| 30 Apr 2025 | ||
| 18:30:16 | ||
| I'll publish a - hopefully good - announcement explaining the situation we're in, and start the ZHF from there. Otherwise time is too limited.. | 20:45:14 | |
| One of the bits of retrospective from 24.11 said that we should start scaling up our builders pretty soon: https://discourse.nixos.org/t/nixos-24-11-retrospective/57085 | 20:54:40 | |
| To my knowledge thats not really possible / helpful because of the limitations of the hydra scheduler | 20:55:32 | |
| I have no horse in this race, but was just the notetaker and that reminded me 😀 | 20:55:41 | |
| (infra, correct me if i'm wrong) | 20:55:46 | |
| current bottleneck is the staging-next of this cycle, which needs K's haskell patches and some darwin love? | 20:57:53 | |
| by the default procedure ZHF would start after this cycle's staging-next is merged. I'm not sure what it really needs, but at least it needs 2-3 more days time, if not more. ZHF can also start before, but then we get breaking changes in the ZHF after it started, so not ideal but acceptable | 21:00:01 | |
| no scaling possible right now | 21:01:04 | |
In reply to @leona:leona.isCool, I'll try to help out with my 128 core system. Will be looking for the ZHF fixes label. | 21:34:50 | |
| 1 May 2025 | ||
| 00:04:56 | ||
| Yeah I'll toss stuff at it for x86_64 on an EPYC, down to help stabilize | 06:02:25 | |
| Have both an EPYC 2 and EPYC 4 to toss at ZHF PRs this time :D | 06:03:37 | |
| Darwin capacity is what hydra is missing | 12:56:48 | |
| We have sufficient Linux capability | 12:57:07 | |
| ZHF started: https://github.com/NixOS/nixpkgs/issues/403336 | 15:11:58 | |
| let's actually mention this in #dev:nixos.org too, then i got almost every place i think :D | 15:12:34 | |
| I've created https://github.com/NixOS/nixpkgs/pull/403348 (in response to the discourse post) This gets the obvious typos out of the way, but I'll revise the documentation in full again to clarify word where/if necessary. Also taking a look at the nixpkgs document now. | 16:16:50 | |
| 20:09:35 | ||