| 19 Feb 2026 |
 emily | shouldn't we be specifying all the features manually? with Meson we disable feature autodetection | 12:37:54 |
| emily | not sure if Qt's build system has something similar we could use | 12:37:58 |
 hexa | Pytest 9 | 12:38:08 |
| hexa | But we keep older versions/hooks around | 12:38:24 |
| emily | no worries, it turned out to just be upstream breakage | 12:38:34 |
| emily | https://github.com/NixOS/nixpkgs/pull/492085, https://github.com/NixOS/nixpkgs/pull/492086 | 12:43:06 |
| emily | I can kick off a qtbase build with https://github.com/NixOS/nixpkgs/pull/492078 if it'd be helpful, otherwise that's probably it for me for the cycle | 12:44:56 |
 vcunat | I've built it on aarch64-darwin already. | 12:46:19 |
| vcunat | Finished just now. | 12:46:24 |
| emily | ah, thank you | 12:46:26 |
| emily | FWIW it'd be zero rebuilds to re-enable it and turn the flag off in Qt instead | 12:46:45 |
| emily | so probably safe to merge even if we take a different tack | 12:46:52 |
| vcunat | Thank you 🙂 | 12:52:35 |
| emily | er, sorry for missing the postPatch there | 13:22:43 |
| emily | I was like "why is this there" and didn't even think about it when doing rebuild avoidance | 13:22:53 |
 Fabián Heredia | Hi, could someone trigger a staging-next eval? linux queues are almost empty and there are a couple of fixes for regressions on staging-next | 15:12:19 |
| Fabián Heredia | * Hi, could someone trigger a staging-next eval? linux queues are almost empty and there are a couple of fixes for regressions on staging-next (and last eval is >24 hrs ago) | 15:12:33 |
| vcunat | Let me queue some merges first. | 15:15:35 |
| emily | do we have https://github.com/NixOS/nixpkgs/pull/490978 on the radar btw? | 15:47:02 |
| emily | currently the cache is going to be broken on -next merge | 15:47:26 |
| vcunat | I know about it. | 15:52:20 |
| vcunat | (watching PRs against the branch) | 15:52:29 |
| vcunat | But I haven't had the energy to review so far. | 15:52:49 |
| vcunat | It's apparently not a big rebuild, at least. | 15:53:03 |
| emily | it's 0-rebuild | 15:53:06 |
| emily | anything that wasn't 0-rebuild would be world-rebuild | 15:53:13 |
| emily | so there's not really any other option for this cycle | 15:53:18 |
| vcunat | I focused on larger rebuilds first. | 15:53:22 |
| emily | (for future cycles it should be changed, of course) | 15:53:25 |
| vcunat | We also have some broken NixOS tests (channel blockers). | 15:53:42 |
