| 26 Apr 2026 |
 Mic92 | @Ericson2314:matrix.org: currently derivations get garbage collected and only a few generations are hold back. If do something similar, we should be good, no? | 17:31:44 |
 John Ericson | well we keep all the old steps | 17:32:00 |
| John Ericson | but yeah we can expire different parts of the data at different times | 17:32:11 |
| John Ericson | BMG: I have some interesting migrations on https://github.com/obsidiansystems/hydra/commits/build-step-by-drv-path this branch | 19:42:44 |
| 27 Apr 2026 |
 BMG | Ok cokl, I'm out today but I'll have a look tomorrow | 06:08:36 |
| BMG | * | 06:08:46 |
|  Ninja joined the room. | 14:37:13 |
| 29 Apr 2026 |
 hexa | https://www.openwall.com/lists/oss-security/2026/04/29/1 | 00:19:34 |
| hexa | starman http request smuggling | 00:19:39 |
 stigo | yep | 00:53:05 |
| stigo | * yep, https://github.com/NixOS/nixpkgs/pull/514601 | 00:54:29 |
| hexa | merged | 07:42:08 |
| John Ericson | BMG: had a look yet? alos still curious about whether any buildsteps are missing a drv path | 18:08:35 |
| John Ericson | if every build step has one, I figure we can send in that migration asap to formally mark the field non-null, even before the new queue runner lands | 18:31:37 |
| BMG | Been tied up with other things. But isn't this just a quick db query against hydra's db? | 18:48:09 |
| BMG | @Ericson2314:matrix.org | 18:48:22 |
| John Ericson | BMG: yup! but I don't have the DB :) | 18:48:43 |
| hexa | bmg has access to the old db host | 18:49:24 |
| hexa | so does simon | 18:49:27 |
| John Ericson | oh cool | 18:49:39 |
| BMG | Ah, @hexa:lossy.network was my goto remote hydra query runner and purveyor of dumps when I was looking into things | 18:49:40 |
| John Ericson | I thought I would have to download many GiB :) | 18:49:47 |
| hexa | root@haumea.nixos.org | 18:49:49 |
| BMG | Oh, I had actually forgotten about that | 18:49:55 |
| BMG | Ah ok lol, I'll poke it first thing in the morning | 18:50:23 |
| John Ericson | if I could get ready only access that would be amazing | 18:50:24 |
| John Ericson | and then I can stop bugging BMG so much :) | 18:50:34 |
| BMG | before I get distracted with other things | 18:50:39 |
| John Ericson | (well actually I would be curious to test out migration time too, but I don't want to fuck up the archival copy) | 18:50:53 |
| John Ericson | BMG: you may also be interested to know I just landed some empty-db migration testing with all the historical schemas | 18:53:05 |
