| 30 Nov 2025 |
Rutile (Commentator2.0) feel free to ping | have a good nom :3 | 18:55:23 |
helle (just a stray cat girl) | good food is important and raclette is way too tasty to care about some CI issues | 18:57:11 |
raitobezarius (DECT: 7248) | After raclette and me rolling down the street, Darwin capacity is temporarily up again | 22:55:40 |
| 1 Dec 2025 |
John Ericson | In reply to @puck:puck.moe well, i mean, i have Opinions on what is represented in a .drv, and really, what the usecase of a .drv is @puck:puck.moe: did you write those up yet? | 01:31:44 |
| @georgyo:nycr.chat left the room. | 02:57:28 |
Rutile (Commentator2.0) feel free to ping | This fucking stupidity has cost me probably ~3-5 hours of stupid debugging time https://github.com/ToruNiina/toml11/issues/303 cc piegames Qyriad | 19:25:23 |
Rutile (Commentator2.0) feel free to ping | https://gerrit.lix.systems/c/lix/+/4676
this refactor is now reviewable | 19:58:51 |
helle (just a stray cat girl) | I want to go over that again when I am doing doc (build) fixes, so I want to check if the code is not going to be a maintenance nightmare for it | 20:02:12 |
helle (just a stray cat girl) | looks good at initial skim for that purpose though | 20:10:00 |
helle (just a stray cat girl) | (not going to claim review on it, because that is only a skim and not looking at other factors then doc-gen really) | 20:11:07 |
just1602 | I know we don't want to format the cpp code, but I run formatter automatically on save and in a CL I just did it reformat a nix file I edit. It makes the diff noisy, but on the other hand I'm wondering if we shouldn't format the nix code since we have an RFC compliant formatter. Or at least is it ok if nix file we change are reformatted?
gerrit should really have a diff view like diffstatic, so you can skip the format change and just focus on the logic. | 21:33:25 |
Rutile (Commentator2.0) feel free to ping | machine configuration via toml is now reviewable :D
https://gerrit.lix.systems/c/lix/+/4533 | 21:52:28 |
Rutile (Commentator2.0) feel free to ping | * machine configuration using a toml file is now reviewable :D
https://gerrit.lix.systems/c/lix/+/4533 | 21:52:36 |
Zoe Z | this had me puzzled for probably longer than it should have https://gerrit.lix.systems/c/lix/+/4680 | 21:58:25 |
just1602 | Is there an easy way to run CI locally? I try to upgrade lix to nixpkgs 25.11 and it always work locally, then a different test from the CI fail when I update my CL :/ | 23:21:34 |
piegames | Yes, the CI jobs are just some flake commands | 23:30:06 |
| 2 Dec 2025 |
raitobezarius (DECT: 7248) | In reply to @mzerodev:matrix.org this had me puzzled for probably longer than it should have https://gerrit.lix.systems/c/lix/+/4680 F | 01:27:52 |
Unhooked Spirits | there's a second typo in the description which confounded me further
typo of aws-cpp-sdk-transfer as aws-cpp-std-tranfer prevents linking
transfer vs tranfer
| 01:32:44 |
just1602 | I've the following error in CI and I'm not sure to understand what's going on:
❯ nix build --no-link .#hydraJobs.tests.nixpkgsLibTests.x86_64-linux
error:
… while evaluating an expression to select 'drvPath' on it
at «internal»:1:552:
… while evaluating strict
at «internal»:1:552:
(stack trace truncated; use '--show-trace' to show the full trace)
error: attribute 'attrpathsSuperset' missing
at /nix/store/2nsijksh4dfi684an1vzig7ilz0hfljc-source/flake.nix:486:77:
485| ++ lib.optionals pkgs.stdenv.isLinux [
486| ((pkgs.callPackage "${nixpkgs}/ci/eval" { inherit nix; }).attrpathsSuperset {
| ^
487| evalSystem = system;
note: trace involved the following derivations:
derivation 'nixpkgs-lib-tests'
| 05:01:21 |
jade_ | i think you can turn off whitespace somewhere | 07:16:59 |
raitobezarius (DECT: 7248) | nixpkgs CI entrypoint moved in 25.11 | 11:28:29 |
just1602 | https://gerrit.lix.systems/c/lix/+/4679
So the issue is I'm trying to bump lix to nixos-25.11-small, two questions:
- nix 2.3 has been removed from nixpkgs, should we removed it too or we still want it?
- the only way I have found to bring back nix_2_3 in the setup is to have a copy of nixpkgs 25.05 because I wasn't able to make it work like 2.18 is setup, I'm wondering if this can cause the issue I'm facing here.
| 13:41:06 |
WeetHet | Flag me if I'm wrong but it's weird that we allow ipc-posix* but don't allow ipc-sysv in the macOS sandbox. I've asked about it in the Nix on macOS channel as well but here's a cl for now:
https://gerrit.lix.systems/c/lix/+/4681
| 18:14:56 |
WeetHet | Oh, there was a PR in cppnix a yeah ago: https://github.com/NixOS/nix/pull/10878 Credit: niklaskorz | 18:30:49 |
WeetHet | Oh, I need to port https://github.com/NixOS/nix/pull/14459 | 18:33:12 |
niklaskorz | here the actual issue about ipc-sysv where it was rejected before: https://git.lix.systems/lix-project/lix/issues/623 | 18:33:45 |
niklaskorz | well technically you can still compile postgres on darwin with posix sempahores instead: https://github.com/postgres/postgres/blob/master/src/template/darwin | 18:43:55 |
niklaskorz | * well technically you can still compile postgres on darwin with posix sempahores instead of sysv: https://github.com/postgres/postgres/blob/master/src/template/darwin | 18:44:24 |
WeetHet | Yeah I guess that would be better... | 19:20:05 |
Rutile (Commentator2.0) feel free to ping | https://gerrit.lix.systems/c/lix/+/3922 can this get abandoned, as it has a -2 (and pollutes my "incoming reviews" box) cc: jade_ | 19:56:32 |