| 30 Jan 2024 |
|  @lab-cat:ag-link.xyz left the room. | 16:32:20 |
|  hexa joined the room. | 22:24:28 |
| hexa | looking into the reproducibility of a package, so I've built it, and building it with --check succeded | 22:25:21 |
| hexa | but that doesn't mean it is reproducible per se, right? | 22:26:41 |
| hexa | but now I only have the one output, and no way to diff it | 22:27:00 |
 atemu12 | It's a good indicator that it's indeed reproducible but there's never a guarantee from any single check | 22:27:38 |
| atemu12 | * It's a good indicator that it's indeed generally reproducible but there's never a guarantee from any single check | 22:27:51 |
| atemu12 | You'd have to check reproducibility on a machine that is as different as possible from the other machine | 22:28:49 |
| hexa | but still the saem architecture | 22:29:49 |
| hexa | * but still the same architecture | 22:29:54 |
| atemu12 | Of course | 22:29:55 |
| atemu12 | A different hostSystem would result in a different drv anyways | 22:30:37 |
| hexa | like I think the website does not really give actionable information beyond running with --check | 22:32:36 |
 raboof | indeed if --check succeeds there is nothing to go on. When it fails and you passed --keep-failed it'll keep the second output, so you can diff it. the reports like https://reproducible.nixos.org/nixos-iso-minimal-runtime/ show that diff formatted with diffoscope | 22:35:03 |
