| 6 Feb 2025 |
 Arian | I think the choice to make a loop device is a userspace one so mount needs to be taught that it can skip that for erofs | 11:31:15 |
 Ramses 🇵🇸 | Right, makes sense | 13:09:56 |
| Arian | Maybe there is already code in master branch for it | 13:12:16 |
| Arian | I'd check. | 13:12:18 |
| Ramses 🇵🇸 | I had a quick look and didn't see any at first sight | 13:14:36 |
|  SomeoneSerge (back on matrix) changed their display name from SomeoneSerge (Gand St. Pieters) to SomeoneSerge (UTC+U[-12,12]). | 17:51:04 |
| 7 Feb 2025 |
 @srestegosaurio:tchncs.de | TIL that you can define drop-ins. Until now, I was overriding everything monkey-style. | 13:09:44 |
| @srestegosaurio:tchncs.de | Nice QoL improvement. | 13:10:06 |
| Arian | https://github.com/NixOS/nixpkgs/pull/379629 | 13:12:04 |
| Arian | ^ could use a review here. Unbreaks TRANSPORT=audit which we silently broke some releases ago | 13:12:21 |
| Arian | but also makes sure we don’t regress on spamming audit logs to console when enabled | 13:12:35 |
| Arian | oh. editorconfig failrue. fun | 13:12:59 |
| Arian | fixed. PTAL | 13:17:16 |
 raitobezarius | cool I had this bug | 13:58:21 |
| raitobezarius | I just never had time to RCA it | 13:58:25 |
| raitobezarius | thanks arian! | 13:58:27 |
| Arian | Container Optimized OS (That thing from Google) has journal audit enabled and auditd disabled but has no kmesg log-sam | 14:56:17 |
| Arian | I have the feeling they might ship a kernel patch for that or something | 14:56:32 |
| Arian | * | 14:59:54 |
| raitobezarius | auditd disabled but journal audit? | 16:54:04 |
| raitobezarius | weird | 16:54:08 |
| Arian | you mean their setup is weird? | 16:56:48 |
| raitobezarius | ye | 16:58:36 |
 ElvishJerricco | Arian: So I'm completely unfamiliar with how this audit stuff works. Is there anything we need to do in stage 1 for this? Like do we need to ensure that stage 1 journald does or doesn't enable auditing or something? | 17:58:09 |
| Arian | I think the whole idea of journald enabling auditing is broken and should be ignored | 17:58:36 |
| Arian | hence defaulting it to not doing anything | 17:58:48 |
| ElvishJerricco | sure | 17:58:54 |
| ElvishJerricco | but if upstream defaults to true, then we need to disable it in stage 1, right? | 17:59:41 |
| Arian | if stage-1 doesn’t have auditing enabled (doesn’t ship auditd; and also journald doesn’t enable it) then the audit logs will just buffer in ak ernel buffer | 17:59:44 |
| Arian | yeh good point. but I don’t think we ship the socket in stage-1 which means the whole functionality is disabled | 18:00:03 |
