| 11 Sep 2025 |
 raitobezarius | if jade_ is fine with it, anyone can take over this CL and pushes it to the finish line | 11:15:20 |
 emily | that's fun. I'd been annoyed by that | 12:45:16 |
 helle (just a stray cat girl) | I need to figure out what is actually breaking, but seeing indefinite hangs on cache.nixos.org downloads at times
though one part of the fix is for me to actually work on implementing the http testing that is badly needed | 12:49:15 |
| raitobezarius | we have USDT tracepoints for curl downloads | 12:52:11 |
| raitobezarius | so you could use DTrace/BPF to show the histograms of sizes passing through the curl thread's buffers | 12:52:28 |
| helle (just a stray cat girl) | that is at least a start for this particular one yeah, any info on this documented? | 12:53:07 |
 jade_ | lix/libstore/trace-probes.d, but there's just the one, it was written up to demo the infra primarily | 17:14:12 |
| 12 Sep 2025 |
|  nightuser joined the room. | 13:30:35 |
 Marie | would it be appropriate to backport this to 2.93?
https://gerrit.lix.systems/c/lix/+/4156 | 16:33:17 |
| raitobezarius | I don't see a big problem personally | 16:36:11 |
| Marie | the meson version bump might be not so nice | 17:31:06 |
| Marie | found a way to work around that | 17:58:57 |
| Marie | ughh i just noticed my check is wrong, it should be the other way around | 19:39:54 |
| emily | raitobezarius: jade_: is there still a need for capnproto-lix on 25.11? looks like all of the debug/coroutine changes are in Nixpkgs, and Nixpkgs is on a newer version with the patches upstreamed | 21:05:58 |
| emily | the old version is broken by CMake 4, so I would at least like to make it conditional on pkgs.capnproto.version | 21:06:15 |
| emily | oh wait, it was removed from main | 21:07:07 |
| emily | I was just trying to build 2.93. ok never mind then | 21:07:15 |
| Marie | is the asanBuild check flaky? it fails on buildkite for https://gerrit.lix.systems/c/lix/+/4160 but it works locally for me | 21:51:11 |
| raitobezarius | ASAN exposes more flakiness of our test suite | 22:19:07 |
 Qyriad | btw is it normal to have thousands of logs of
Yuki pasta[1161683]: lseek() failed on /proc/net file
Yuki pasta[1161683]: : Illegal seek
| 22:41:18 |
| Qyriad | (in journald from nix-daemon.service) | 22:42:07 |
 Lun | Is that https://lore.kernel.org/all/20250815195616.64497967@chagall.paradoxon.rec/ ? | 22:44:41 |
| Qyriad | we are on 6.12.42 | 22:46:20 |
| Qyriad | so maybe | 22:46:22 |
| raitobezarius | In reply to @qyriad:katesiria.org
so maybe This is it | 23:31:20 |
| raitobezarius | Fixed in latest stables | 23:31:34 |
| 13 Sep 2025 |
| emily | how much of a pain would it be to get a 2.32 release cut with the toml11 fixes? | 03:49:24 |
| emily | er. 2.93 | 03:49:31 |
 aloisw | It was not removed, just moved to the overlay to fix the splicing. | 05:18:55 |
| emily | so it was. I wonder why the build isn't broken with it then. | 05:19:39 |
