| 1 May 2024 |
 hexa | thanks 🙂 | 17:08:34 |
 adamcstephens | i suspect that was a broader category than just the LXC related work :) | 17:16:27 |
| 2 May 2024 |
| adamcstephens | ok, i did remove the testing refactor from https://github.com/NixOS/nixpkgs/pull/307039 | 14:09:41 |
| adamcstephens | if nothing else, some input on the name of the option is desired | 14:10:08 |
| 4 May 2024 |
 rungmc | Just roll with that. Communicates what the option is pretty well. I feel like something like that softReload suggestion would EVENTUALLY have me looking it up down the line to see why I set it in my config. | 18:13:17 |
| rungmc | Is it an accurate read of what's going on here to say that if you were to set the option and alter a profile in the preseed that the affected containers would update individually on their restart? Trying to work through how I'd want to mesh that with autoUpgrade. | 18:24:30 |
| adamcstephens | I actually don’t know what happens if you try and modify a profile that running containers use. I’d have to test that. | 18:59:49 |
| rungmc | I think Incus typically tries to make changes immediately. | 19:16:19 |
| adamcstephens | K. Anyway I don’t think this change affects changing preseed. That’s already a standalone service unit | 19:29:52 |
| adamcstephens | This more about not shutting down instances when the incus service is stopped. Mostly the benefit is allowing the host incus to be updated without downtime to instances | 19:30:36 |
| rungmc | Fair. I suppose I'm barking up the wrong tree a bit trying to figure out if there's a trade-off there that I care about. | 19:36:14 |
| adamcstephens | this is the default setup when using the zabbly packages, and probably on other distros. i carried our setup forward from lxd, but it finally annoyed me enough to implement it as an option | 19:39:06 |
| adamcstephens | personally i would rather the instances stay up when something triggers the incus service to require a restart. for example, any package dependencies initiating a rebuild will do so. | 19:40:07 |
| adamcstephens | i run unstable everywhere right now, so it's a bit painful to take an outage of 11 instances just to do updates on the host :) | 19:45:27 |
| adamcstephens | * i run unstable everywhere right now, so it's a bit painful to frequently take an outage of 11 instances just to do updates on the host :) | 19:47:57 |
| rungmc | Same. I also ran LXD/Incus on Debian for a while before going all-in on NixOS, so I think you just cleared up a big disconnect for me. | 19:48:00 |
| adamcstephens | i could probably be convinced to make this the default. technically there's some minor loss of reproducibility, but maybe it's really not a problem | 19:49:09 |
| rungmc | It's probably fine. I would imagine there's a limit to how reproducible Incus can really be without encroaching on its features. I have no idea how you'd ever set it up so you could define an instance in Nix and make that work with clustering, for example. | 19:53:51 |
| rungmc | Using a NixOS container image and having the flexibility to just lift and shift between servers is good enough for me. | 19:57:22 |
| adamcstephens | yeah i sometimes overthink these things :) | 20:13:07 |
| rungmc | Speaking of overthinking things, Openwrt on Incus on NixOS works brilliantly. The devices I was working with are still terrible but at least I verified that their version of hostapd is huge improvement 😉 | 20:18:40 |
| 6 May 2024 |
| adamcstephens | one thing i've long wanted is a way to nix copy into an instancwe | 19:34:41 |
| adamcstephens | * one thing i've long wanted is a way to nix copy into an instance | 19:34:42 |
| adamcstephens | e.g. `nix copy .#default --to incus://myinstance | 19:35:34 |
| adamcstephens | * e.g. `nix copy .#default --to incus://myinstance` | 19:35:37 |
| adamcstephens | element why, just why are you so dumb | 19:35:51 |
| 7 May 2024 |
| adamcstephens | hexa: is your incus poc(?) still going on? | 18:33:27 |
| adamcstephens | I'm curious if you use the web UI | 18:33:37 |
| hexa | yeah, I've dabbled a bit with the web ui | 18:35:14 |
| hexa | but I'm reworking the deployment currently, so not accessible | 18:35:25 |
