| 1 Jul 2024 |
 Mic92 | CVE-2024-6387 | 12:00:11 |
 emily | note that that replaces an RCE with no currently-known exploit for NixOS with I think an unauthenticated DoS | 14:35:49 |
| emily | probably the right tradeoff though | 14:35:58 |
| emily | hopefully it'll be obsolete in a few hours :) | 14:36:18 |
|  Olaf Krasicki Freund joined the room. | 21:57:01 |
| 2 Jul 2024 |
 mei 🌒& | Mic92: please remove me from nix-doom-emacs maintainers (team/repo desc)
https://github.com/NixOS/nixpkgs/pull/323895 | 00:20:49 |
| Mic92 | Done | 04:10:09 |
| Mic92 | We should probably just archive it given it's disfunctional state. | 04:11:34 |
| Mic92 | I will open an issue | 04:12:04 |
|  Zyansheep joined the room. | 04:31:18 |
|  @janik0:matrix.org left the room. | 13:51:16 |
| 3 Jul 2024 |
|  nora joined the room. | 05:30:31 |
 Matt Sturgeon | Is buildbot having issues atm? Doesn't seem to be picking up PR events and the web portal is timing out for me | 15:48:16 |
| Matt Sturgeon | * Is buildbot having issues atm? Doesn't seem to be picking up PR events and the web portal is timing out for me
EDIT: Seems resolved | 16:12:48 |
| Matt Sturgeon | * Is buildbot having issues atm? Doesn't seem to be picking up PR events and the web portal is timing out for me
EDIT: Seems resolved? | 16:56:34 |
 zowoq | In reply to @mattsturg:matrix.org
Is buildbot having issues atm? Doesn't seem to be picking up PR events and the web portal is timing out for me
EDIT: Seems resolved? Looks like it runs out of memory and sometimes reboots while trying to evaluate nixvim. cc Mic92 | 23:37:15 |
| 4 Jul 2024 |
|  @3650badcop:catgirl.cloud changed their profile picture. | 01:39:47 |
| Mic92 | Could it be that it true | 03:34:39 |
| Mic92 | Could it be that it also evaluates something else in addition in the background? | 03:35:46 |
| Mic92 | I have quite a few bigger projects on machine with less RAM but it's stable for me so far. But I only have one ci running | 03:36:51 |
| Mic92 | If it goes out of memory there should be also a dmesg log event that contains all the processes at the time | 03:38:33 |
| Mic92 | That would be useful. Usually Linux also wouldn't usually crash from that. | 03:39:10 |
| Mic92 | Is it by chance happening on an hetzner ax102? We had lately also crashes with other types of load on that instance type. Garnix reports similar things | 03:41:02 |
| zowoq | Nothing in the background, just nixvim. I haven't seen it do this before, usually multiple buildbot evals are fine, even with hercules/hydra running as well. It isn't an ax102, build03 is a ryzen 9 3900, 128gb. | 04:20:48 |
| Matt Sturgeon | If you're looking at builds for this PR it might be that the pr itself is fundamentally flawed.
It has a submodule option that is essentially doing submoduleWith { modules = [ ../. ]; } so that's probably blowing up the module system by recursively including itself in the submodule. | 04:35:16 |
| Matt Sturgeon | In reply to @mattsturg:matrix.org
Is buildbot having issues atm? Doesn't seem to be picking up PR events and the web portal is timing out for me
EDIT: Seems resolved? Idk what was causing this (seems resolved on other branches). The last few build failures are probably red herrings | 04:37:41 |
| zowoq | Just tried nixvim main branch and it rebooted. | 04:52:31 |
 Gaétan Lepage | In reply to @magic_rb:matrix.redalder.org
If i find time, (procrastinate studying for exams) i might give it a quick peek to see if i can fix it quickly. It seems like a productivity blocker and those we dont want The ability to manually stop/cancel builds would let us partially unload the builder as much as possible. | 08:25:54 |
| Gaétan Lepage | Build-bot doesn't cancel a build if the PR branch gets updated.
Hence, pushing 10 times on a branch would start 10 full CI runs (which are quite expensive). | 08:26:47 |
 magic_rb | In reply to @glepage:matrix.org
The ability to manually stop/cancel builds would let us partially unload the builder as much as possible. I will look into that today, the permission system needs updating | 08:40:23 |
