| 20 Nov 2025 |
 vcunat | * staging-next will most likely get merged this weekend. | 23:09:42 |
| 21 Nov 2025 |
 K900 | @vcunat so uh | 09:25:56 |
| K900 | GTK4 update | 09:25:59 |
| K900 | lol | 09:25:59 |
| vcunat | What do you mean? | 09:26:31 |
| K900 | The Vulkan corruption is apparently caused by them holding it absolutely mind-numbingly wrong | 09:26:33 |
| K900 | There's a GTK 4.20.3 | 09:26:43 |
| vcunat | We need to rebuild gtk4 to fix this? | 09:26:44 |
| K900 | That has a fix | 09:26:46 |
| K900 | We COULD | 09:26:50 |
| K900 | And the blast radius is not horrendous | 09:27:03 |
| K900 | And Mesa can't fully fix this | 09:27:15 |
| K900 | Because again they are doing things that are juggling loaded footguns levels of unsafe | 09:27:32 |
| vcunat | So we at least open a PR to see the rebuild estimate? | 09:29:13 |
| vcunat | Currently we're mostly waiting for darwin. | 09:29:23 |
| K900 | nix-drv-index on a somewhat old database says 2k per Linux | 09:29:43 |
| vcunat | So if it causes no big darwin rebuild, it might be OK for release. | 09:29:49 |
| vcunat | * So if it causes no big darwin rebuild, it might be OK for release schedule. | 09:29:58 |
| vcunat | Are you working on a PR or should I? | 09:30:56 |
| K900 | On it | 09:31:11 |
 uep | yeah, it's a mess | 09:32:10 |
 leona | yes, I'd say that that's fine | 09:32:22 |
| uep | @k900: happy to try that patch, or your PR.. | 09:34:46 |
| uep | * K900: happy to try that patch, or your PR.. | 09:35:00 |
| K900 | https://github.com/NixOS/nixpkgs/pull/463695 | 09:35:02 |
| K900 | Can you try the GTK bump? | 09:35:07 |
| uep | the patch seems to be draft so presumably not the sme as whatever's in the bumped release? | 09:35:30 |
| K900 | The patch is not on GTK | 09:36:04 |
| K900 | It's a workaround in Mesa | 09:36:09 |
| vcunat | I suppose we want (to test) the GTK bump with unpatched mesa 25.3.0 | 09:37:56 |
