| 2 Jul 2026 |
 hexa | https://github.com/search?q=repo%3Acalmjs%2Fcalmjs%20pkg_resources&type=code | 15:05:18 |
| hexa | what kind of runtime setuptools legacy garbage is this | 15:05:26 |
| hexa | we absolutely cannot propagate setuptools_80 at runtime | 15:05:51 |
| hexa | https://github.com/calmjs/calmjs/issues/70 | 15:06:28 |
 emily | oh ffs | 15:12:59 |
| emily | I think ffmpeg will still be broken for ld reasons | 15:13:03 |
| hexa | what a fun cycle | 15:13:17 |
| emily | maybe we should have just restarted the cycle for that | 15:13:29 |
| emily | given we ended up doing the libssh2 thing anyway | 15:13:34 |
| hexa | wdyhm | 15:13:41 |
| hexa | Redacted or Malformed Event | 15:13:43 |
| hexa | merge staging again? | 15:13:50 |
| emily | just reverting the ld64 issue the first time it came up | 15:13:53 |
| emily | and rerolling Darwin stdenv | 15:13:58 |
| emily | since we threw away so many builds for libssh2 it might not have been that significant an extra delay | 15:14:17 |
| emily | https://github.com/NixOS/nixpkgs/pull/537840 would appreciate it if someone could test, don't have the time to run a build right now, should be doing other things 🫠| 15:16:21 |
 Myria | Yeah, swift-wrapper ignores the -fuse-ld passed to it, so a dependency of swiftpm fails. | 15:20:29 |
| Myria | Building now, will let you know | 15:25:40 |
 Sergei Zimmerman (xokdvium) | WFM | 15:29:41 |
| emily | merged, thank you | 16:09:13 |
| emily | Myria: is there an open PR for the SDL thing? | 16:10:11 |
 Grimmauld (any/all) | nothing that pinged SDL team i think | 16:14:47 |
| Myria | No, just mentioned it because I ran into it | 16:52:19 |
| emily | am testing a build of it locally | 16:52:22 |
| Myria | Too easy | 16:52:29 |
| emily | I hope it's not the case that everything downstream of a failure needs the fix too | 16:52:43 |
| emily | or we're screwed | 16:52:46 |
| Myria | Yosys downstream of gtk2 was fine, so it isn't a certain thing | 16:54:02 |
| emily | https://github.com/NixOS/nixpkgs/pull/537877 | 17:13:41 |
| emily | should be enough to build up to full FFmpeg | 17:14:00 |
