| 21 Sep 2025 |
K900 | Yay 10 year old JSON parser | 08:13:44 |
K900 | Very barely more alive fork at https://github.com/containers/yajl | 08:15:42 |
K900 | @emily thoughts? | 08:17:57 |
K900 | (I know you enjoy these kinds of things) | 08:18:30 |
K900 | The funny thing is, the fork doesn't even change | 08:23:48 |
K900 | * The funny thing is, the fork doesn't even change the cmake version | 08:23:51 |
K900 | But it does have other fixes we probably want | 08:23:58 |
ElvishJerricco | actually, I guess I could PR the patch that systemd initrd needs right now if that means we can get it in staging-next today, which would let me finish off systemd-initrd-by-default over the next days. Does that work? | 08:25:32 |
K900 | I think we should hold off | 08:25:50 |
K900 | This cycle is already very cursed | 08:25:56 |
ElvishJerricco | ah alright | 08:26:01 |
K900 | We have cmake 4 and LLVM 2 | 08:26:08 |
K900 | * We have cmake 4 and LLVM 21 | 08:26:09 |
K900 | I would REALLY like to not add a systemd update into the mix | 08:26:24 |
ElvishJerricco | the benefit of doing it this cycle is just that systemd initrd would benefit from more people using it sooner before release, but it's not that big a deal | 08:26:37 |
ElvishJerricco | (well, the systemd "update" would just be a small patch to the shutdown code) | 08:27:13 |
ElvishJerricco | (i'm not talking about 258 right now) | 08:27:25 |
K900 | Oh | 08:27:29 |
K900 | Ehhh | 08:27:35 |
ElvishJerricco | still just don't want to introduce systemd-initrd-by-default in this cycle? That's fine | 08:28:28 |
K900 | I am at this point genuinely just afraid of touching it lol | 08:28:54 |
K900 | (outside of fixing cmake 4) | 08:28:59 |
ElvishJerricco | lol ok | 08:29:30 |
ofalvai | Small fix for toml11 and LLVM 21: https://github.com/NixOS/nixpkgs/pull/444901 | 09:03:41 |
K900 | OK there's a lot | 09:54:16 |
K900 | Of cmake 4 fails | 09:54:24 |
K900 | And I'm not even close to done | 09:54:34 |
Vladimír Čunát | Should we call an early ZHF for staging-next this time? 😁 | 11:03:04 |
Grimmauld (any/all) | there is still libxml that'll break shit | 11:46:28 |
Vladimír Čunát | I don't expect we even need it in 25.11 when it comes to it. | 11:47:26 |