| 3 Nov 2021 |
 das_j | I miss the conventional way of packaging where not everything is being rebuilt and people find random problems over the next months :/ | 18:53:33 |
| das_j | * I miss the conventional way of packaging where not everything is being rebuilt and people find random problems over the next months. Then it's not the problem of the coreutils/glibc bumper | 18:53:49 |
 moritz.hedtke | In reply to @janne.hess:helsinki-systems.de
I miss the conventional way of packaging where not everything is being rebuilt and people find random problems over the next months. Then it's not the problem of the coreutils/glibc bumper Yeah way better (imagine loosing tons of data because if this issue) | 18:54:58 |
| das_j | I'll start by imagining using ZFS ;) | 18:55:42 |
| das_j | * I'll start by imagining using ZFS ;) one step at a time | 18:55:49 |
| moritz.hedtke | In reply to @janne.hess:helsinki-systems.de
I'll start by imagining using ZFS ;) one step at a time Well doesn't matter if your build servers use zfs ^^ | 18:57:26 |
| moritz.hedtke | (and you use the provided binaries) | 18:57:51 |
 Vladimír Čunát | That also contributed to confusing the issue. Some builds succeeded on Hydra but they had broken output, and you were getting them substitute as dependencies. | 19:07:28 |
| Vladimír Čunát | * That also contributed to confusing the issue. Some builds succeeded on Hydra but they had broken output, and you were getting them substituted as dependencies. | 19:07:32 |
| Vladimír Čunát | In reply to @janne.hess:helsinki-systems.de
I miss the conventional way of packaging where not everything is being rebuilt and people find random problems over the next months. Then it's not the problem of the coreutils/glibc bumper That would make the problem even harder overall. | 19:08:32 |
| Vladimír Čunát | In reply to @janne.hess:helsinki-systems.de
I miss the conventional way of packaging where not everything is being rebuilt and people find random problems over the next months. Then it's not the problem of the coreutils/glibc bumper * That would make the problem even harder overall. (though perhaps not for you :-) | 19:08:56 |
| moritz.hedtke | I was like: I have a bit of time and investigating something larger and more complex is probably fun - well it was not :D but still worth it - I learned a lot | 19:09:01 |
| 4 Nov 2021 |
 trofi | firefox stopped compiling in staging. Bisect is not very helpful and point at merge commit: https://github.com/NixOS/nixpkgs/commit/703ffa2f5d1d7b4d1dc60511be8badc0f52e0b26
Last few lines of build failure: https://dpaste.com/GTEHY9Z94.txt. As if a patch is missing or clang version changed. I think it's a bug in libcxx's headers. But it's slightly surprising to see a failure at merge time.
| 12:09:32 |
 Sandro | It is not related to https://github.com/NixOS/nixpkgs/commit/703ffa2f5d1d7b4d1dc60511be8badc0f52e0b26#diff-8fee7335b98c5c1dd27e3a6673fa0b3ed41ff6a1c754b1c5491392e34b0ff259R36? | 12:49:23 |
| Sandro | * It is not related to https://github.com/NixOS/nixpkgs/commit/703ffa2f5d1d7b4d1dc60511be8badc0f52e0b26#diff-8fee7335b98c5c1dd27e3a6673fa0b3ed41ff6a1c754b1c5491392e34b0ff259R36 ? | 17:27:22 |
| trofi | Oh, right. git show 703ffa2f5d1d7b4d1dc60511be8badc0f52e0b26 also shows only conflict resolution. Don't know how I ran it before when I saw nothing in conflicts. | 22:50:06 |
| trofi | Filed https://github.com/NixOS/nixpkgs/issues/144670 | 23:00:13 |
| 5 Nov 2021 |
| trofi | staging-next fails to merge to staging on nixos/tests/kexec.nix: https://dpaste.com/6PEWRL43Y.txt | 08:59:17 |
| Vladimír Čunát | I opened another iteration. I'm not sure about exact timing of merges from staging and staging-next wrt. the ZHF/21.11 schedule, but Hydra wasn't too busy now and I figured we could at least do the builds. | 10:33:31 |
 hexa | Alyssa Ross: I propose we downgrade llvm used in rust from 13->12 until after the release, respectively the next firefox release | 10:49:43 |
| Sandro | That shouldn't have an impact on vendor hashes, right? | 22:33:36 |
|  legendofmiracles joined the room. | 23:56:37 |
| 6 Nov 2021 |
 fabianhjr | kexec-tools build fails on a recent rebase of staging-next to the main branch. Bumping the version (2.0.22 → 2.0.23) does build, almost done with a nixos-rebuild. | 06:04:00 |
| fabianhjr | Just booted into staging-next :3
Everything seems to be in order | 07:27:27 |
|  Tseb joined the room. | 09:19:20 |
| Tseb left the room. | 09:22:23 |
| Tseb joined the room. | 09:24:30 |
| Tseb left the room. | 09:25:26 |
| 7 Nov 2021 |
 Ryan Burns | Would appreciate a quick look at https://github.com/NixOS/nixpkgs/pull/144941, I think that's the only major blocker on x86_64-darwin | 21:07:53 |
| Sandro | merged that | 21:15:52 |
