| 10 Jul 2023 |
 raitobezarius | our protocol to assert the state of the machine is suboptimal for that | 15:01:43 |
 nikstur | wait_for_reboot() or something | 15:01:44 |
| raitobezarius | I think you need to do adjustments for that | 15:01:49 |
| raitobezarius | in the internal state machine | 15:01:53 |
| nikstur | Had hoped, I wouldn't need to touch it | 15:06:43 |
| nikstur | mhhhm | 15:06:49 |
| nikstur | Ok so one thing I figured out is that I need to set machine.connected = False so that it reconnects to the shell | 15:12:23 |
| nikstur | Now I can machine.sleep but thats obv not a very robus solution | 15:12:50 |
| raitobezarius | There was discussion in NixOS Systemd about RIIR the protocol between VM and driver | 15:16:52 |
| raitobezarius | so it would have proper framing, et. | 15:16:56 |
| raitobezarius | * so it would have proper framing, etc. | 15:16:58 |
| nikstur | I read that thread but I don't think this would really improve the situation for waiting for a reboot | 15:56:05 |
| 22 Jul 2023 |
| raitobezarius | nikstur: I think something is very wrong with our NixOS tests wrt to boot | 21:42:34 |
| raitobezarius | And reason for why it's breaking also Lanzaboote tests | 21:42:42 |
| raitobezarius | but I feel like this is driver specific | 21:43:20 |
| raitobezarius | ok it's potentially layer 8 but we need to fix timeouts | 21:59:13 |
| raitobezarius | if you try to boot QEMU test with no media nothing, it will just loop infinitely | 21:59:33 |
