| 4 Jul 2026 |
 hexa | ah yeah, darwin is bored | 17:22:25 |
 emily | I didn't do it :) | 17:24:48 |
| emily | https://github.com/NixOS/nixpkgs/pull/538441 | 17:25:33 |
| hexa | eval killed | 17:27:20 |
| emily | https://github.com/NixOS/nixpkgs/pull/538418 is still waiting on K900, might make sense to eval after that | 17:31:02 |
| emily | testing Swift fix now | 17:31:21 |
| hexa | https://github.com/NixOS/nixpkgs/pull/538442 looking at the awscli2 fix over here | 17:32:15 |
 K900 | In reply to @emilazy:matrix.org
https://github.com/NixOS/nixpkgs/pull/538418 is still waiting on K900, might make sense to eval after that Building | 17:33:21 |
| emily | https://github.com/NixOS/nixpkgs/pull/538444 | 17:35:23 |
| emily | ofc the Swift driver is broken too… | 17:36:47 |
| emily | will add more commmits | 17:36:48 |
 Randy Eckenrode | Hopefully the Swift pain will be over by the next staging-next cycle. | 17:41:59 |
| K900 | I looked away and it died to SDL test flakiness | 17:44:25 |
| emily | at least we're getting to test a lot of things using LLD 🫠 | 17:49:21 |
 Grimmauld (any/all) | wait what? SDL tests shouldn't be flaky at all | 17:59:02 |
| K900 | testrwlock or whatever that WA s | 17:59:32 |
| K900 | * testrwlock or whatever that was | 17:59:38 |
| Grimmauld (any/all) | occasionally pygame breaks to sdl updates because they use weird internal api for stuff, but the tests of sdl itself should be solid | 17:59:39 |
| K900 | I've seen it flake under heavy load | 17:59:46 |
| Grimmauld (any/all) | https://github.com/NixOS/nixpkgs/pull/526570 should be in the cycle, no? | 18:00:05 |
| K900 | Uhhh idk still happens here | 18:00:56 |
| hexa |
Merged via the queue into NixOS:staging with commit e58bb92 on Jun 1
| 18:00:57 |
| hexa | git branch --contains e58bb92 fwiw | 18:01:06 |
| emily | why does swiftpm do its own linking >_< | 18:01:12 |
| hexa | (yes, it is in the cycle) | 18:01:19 |
| Grimmauld (any/all) | if that is in the cycle, do you have a log @K900 ? 300 seconds really REALLY ought to be enough... | 18:01:57 |
| K900 | Uhh I already restarted it | 18:02:16 |
| Grimmauld (any/all) | eh | 18:02:23 |
| K900 | It was concurrent with two webkits though | 18:04:27 |
| K900 | So I think that's very possible | 18:04:31 |
