| 12 Nov 2025 |
 vcunat | I think we just merge staging-next today? | 15:07:13 |
| vcunat | The "cachix on darwin" problem hasn't been resolved yet; I haven't even seen any reaction from the relevant pinged people. But nixos-unstable is linux-only, so I think we can proceed with that one. nixpkgs-unstable would be blocked, but I'd hope that within another day or two the problem would get resolved somehow (while Hydra processes the aftermath of the merge). | 15:07:16 |
 leona | IMHO please, yes. | 15:09:31 |
 Wolfgang Walther | I agree on not blocking on it given there has been no reaction. We'll not be affected by it in CI, because we're not using cachix from Nixpkgs. | 15:09:54 |
| vcunat | staging-next + staging-nixos merged a while ago. | 18:22:26 |
 Fabián Heredia | As a heads up for current nixos unstable/staging-next: https://github.com/NixOS/nixpkgs/pull/458944
Versions 2.4.126 and 2.4.127 contain an ABI breakage in libdrm_amdgpu.
Depending on which Mesa version or commit you use, you may see stack
smashing resulting in crashes of AMD Mesa drivers. It's recommended to never package versions 2.4.126 and 2.4.127.
libdrm 2.4.127 is currently on unstable+staging-next
| 20:21:08 |
| Fabián Heredia | 25.05 is currently on 2.4.124, going to backport all updates | 20:23:10 |
 K900 | Absolutely do not | 20:23:32 |
| Fabián Heredia | :O | 20:23:43 |
| K900 | We may not even have a staging-25.05 | 20:23:47 |
| K900 | And backporting 6 months of libdrm updates will definitely not go well | 20:23:57 |
| Fabián Heredia | Well, yeah thats a good point. I thought stable releases had 1 month of support post-next-stable release | 20:24:34 |
| vcunat | Isn't .124 OK? Or do I read those citations wrong? | 20:41:23 |
 niklaskorz | no you're reading correctly, nothing to do for 25.05 | 20:43:17 |
| Fabián Heredia | yeah, that one is fine | 20:49:34 |
| 13 Nov 2025 |
 ivy | this is so odd im looking at cachix stuff | 06:40:48 |
| ivy | and that export does exist | 06:40:52 |
| ivy | nm libnixexpr.dylib | grep -i __ZN3nix14DerivedPathMapINSt3__13setINS1_12basic_stringIcNS1_11char_traitsIcEENS1_9allocatorIcEEEENS1_4lessIS8_EENS6_IS8_EEEEE10ensureSlotERKNS_17SingleDerivedPathE
U __ZN3nix14DerivedPathMapINSt3__13setINS1_12basic_stringIcNS1_11char_traitsIcEENS1_9allocatorIcEEEENS1_4lessIS8_EENS6_IS8_EEEEE10ensureSlotERKNS_17SingleDerivedPathE | 06:41:11 |
| ivy | like there it is | 06:41:14 |
| leona | Could you try if it builds, if you revert https://github.com/NixOS/nixpkgs/pull/447364 (or some parts of it?). Bisect lead me in this direction, but too much stuff (ghc) didn’t build in between, so it didn’t finish and then I had not the time to test with that reverted | 07:56:26 |
| K900 | So do we have anything left for next cycle | 11:05:23 |
 Grimmauld (any/all) | https://github.com/NixOS/nixpkgs/pull/459474 might make sense | 11:06:07 |
| K900 | The queue will probably drain by tonight | 11:06:15 |
| Grimmauld (any/all) | but i don't know shit about bluez | 11:06:23 |
| K900 | Do we have bluez proper updated | 11:06:47 |
| Grimmauld (any/all) | bluez-proper reuses the source fetcher iirc | 11:07:00 |
| Grimmauld (any/all) | so this is a bluez update | 11:07:31 |
| ivy | i ran out of space on my computer so couldn’t | 11:07:58 |
| niklaskorz | indeed: https://github.com/NixOS/nixpkgs/blob/2a1f3303bfc736cdb23bc593ff72079d82b5a589/pkgs/by-name/bl/bluez/package.nix#L32 | 11:08:21 |
| K900 | OK cool | 11:08:41 |
