| 10 Apr 2026 |
|  @ktrompfl:matrix.org joined the room. | 11:54:28 |
 emily | not sure that's better :D | 12:15:43 |
| emily | systemd stores env var style metadata about cgroups or such right? I guess the full drvPath could go in there? | 12:16:22 |
 raitobezarius | This doesn't ring a bell to me, what are you thinking of? | 12:16:45 |
| emily | I may be completely hallucinating a memory | 12:17:41 |
| raitobezarius | I think I know systemd stores things like INVOCATION_ID and so on, but I don't really know where they live, I think they're in the sd PID1 memory or something | 12:20:17 |
| raitobezarius | I know there's some desires to add xattrs to cgroups inodes to store more data, in relation with the Flatpak sandboxing work and real-time permission bestow/removal | 12:20:46 |
| raitobezarius | But the kernel feature is not even ready yet | 12:20:50 |
| emily | yeah I was thinking more attaching stuff to systemd scope units | 12:22:18 |
| raitobezarius | There might be something but my brain cannot come up with a hit on this sort of question | 12:23:15 |
| @ktrompfl:matrix.org | Hey, I'm not sure if this is the right place to ask, but I have a question about a rust package (https://github.com/Ktrompfl/jay/tree/nix-flake) which builds fine with CppNix 2.31.3 but fails to build with Lix 2.94.1 due to some failing tests.
These tests fail on creating an io_uring instance with os error 38, most likely due to the sandboxed build environment.
Is this intended and the package should just skip the respective tests or is this a problem with either CppNix or Lix? | 12:30:27 |
|  Sergei Zimmerman (xokdvium) joined the room. | 12:31:36 |
 K900 | Yes, intended | 12:32:44 |
| K900 | io_uring is banned in the sandbox | 12:32:49 |
| K900 | Because it's currently not practical to actually sandbox effectively | 12:32:56 |
| @ktrompfl:matrix.org | Ok, thank you! | 12:34:42 |
 sterni | does anyone have a guess what is going on here? https://buildkite.com/afnix/lix/builds/1124#019d767e-d754-4569-852c-51bfa85ce70c | 14:02:50 |
 Yureka (she/her) | I bisected it to the staging cycle which updated perl to 5.42 | 14:03:32 |
| Yureka (she/her) | and opened a topic in the darwin zulip: https://zulip.lix.systems/#narrow/channel/27-Darwin/topic/perl.20installCheckPhase.20fails.20after.20nixpkgs.20update/with/9170 | 14:03:54 |
| sterni | ty! | 14:08:21 |
| sterni | I think next lowdown bump I'm going to just wait with changing the flake until the desired version is default in the channel this is still such a battle with a million annoying little issues. | 14:09:38 |
| Yureka (she/her) | can also bisect further into perl once I'm home ig, but it's not fun because the perl bump required a bunch of nixpkgs changes | 14:10:33 |
| Yureka (she/her) | would be much nicer if someone knew the solution already ;) | 14:11:05 |
| raitobezarius | Sergei Zimmerman (xokdvium): you guys are also affected with the perl bindings right? | 14:34:40 |
| Sergei Zimmerman (xokdvium) | In reply to @raitobezarius:matrix.org
Sergei Zimmerman (xokdvium): you guys are also affected with the perl bindings right? Haven’t seen the issues yet, probably I guess? I have -5% of a clue about Perl | 14:36:04 |
 Display Name | May I ask is there guide to lix?, I have done the obsession that I wanted to do, can I ask is iirc lix is package manager, does it have compilers as a package (or that not how it works), may I ask does it work on alphine linux (for my clunky container, single user rather than multi user) (or that not how its work)? | 16:20:22 |
| K900 | This is a lot of questions | 16:25:14 |
| K900 | Lix is an implementation of Nix, which is kind of a weird meta-buildsystem/pacakge manager thing | 16:25:32 |
| K900 | Lix is an implementation of Nix, which is kind of a weird meta-buildsystem/package manager thing | 16:25:36 |
| K900 | That kind of exists in its own class | 16:25:43 |
