| 28 Nov 2025 |
 aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | https://hydra.nixos.org/eval/1820546#tabs-still-fail | 08:57:54 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | I guess I'm not going to click that expand button several times | 08:58:11 |
 vcunat | I meant that you can crawl Hydra the same way as nix-review-tools. | 08:58:57 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | ...okay | 09:01:59 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | I hope I don't need to get around anubis | 09:02:32 |
| vcunat | (perhaps others can suggest, too) | 09:02:34 |
| vcunat | * (perhaps other people can suggest, too) | 09:02:38 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | So who's managing the hydra instance now? | 09:03:57 |
| vcunat | The infra-build team has access to do that (including myself). | 09:07:09 |
| vcunat | I don't think we're using those. | 09:08:02 |
 Janne | You can look at zhf.fail, I just parse the DOM | 09:08:02 |
| vcunat | IIRC some things were missing the API and the workaround was like that. | 09:08:33 |
| Janne | Maybe I should put the cache of that in s3 so people can access the data without adding more load to hydra 🤔 | 09:08:54 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | In reply to @janne:hess.ooo
You can look at zhf.fail, I just parse the DOM 🫠 | 09:08:58 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | Okay | 09:09:02 |
| vcunat | Well, it's one of the reasons why I suggested to check for presence of log and presence of binaries. | 09:09:55 |
| vcunat | (neither hit hydra.nixos.org) | 09:10:45 |
 leona | oh and crawling hydra is very much not fun. I tried doing it for finding broken builds with something similar to gist.github.com/Mindavi/6b33251a3a7a207e95d0ca289da312cb. And I had to go down to <10 simultaneous connections to not break hydra | 09:12:01 |
| leona | and even then it struggled so much that I stopped it | 09:12:11 |
 Julien | I think for zh.fail we have scripts that mostly work | 09:13:01 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | Why are we all trying to solve a problem that should not exist | 09:13:10 |
| leona | because hydra is an awful, strange perl software that nobody wants to touch and change | 09:13:52 |
| vcunat | Why is everyone solving it from scratch? | 09:13:54 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | And zh.fail doesn't carry drvHash because drvHash requires you to check every page of builds | 09:14:22 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | Which could add even more load to hydra | 09:14:31 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | At the same time, the jump link from nixos-search to hydra is still broken for darwin. | 09:17:43 |
| Janne | Actually… | 09:23:27 |
| Janne | @simon.hauser:helsinki-systems.de 👋 | 09:23:35 |
| leona | could someone change stableBranch / officialRelease to true for nixos/release-25.11, nixos/release-25.11-small, nixpkgs/nixpkgs-25.11-darwin | 12:32:37 |
| leona | as a preparation for the release on sunday | 12:32:49 |
