| 14 Mar 2023 |
 Arson Copperflame | In reply to @k900:0upti.me
You could also set them manually, but generally it might be a better idea to run unstable, as that's what nixos-wsl is mostly tested against Is it? All the tests run against stable and that's what I'm running as well. It's probably true for anything related to nativeSystemd though, because we can't run the tests against that (yet) | 16:26:50 |
 Sandro π§ | We should disable udev again. It draws in a lot https://github.com/NixOS/nixpkgs/blob/nixos-unstable/nixos/modules/services/hardware/udev.nix#L355-L455 | 16:27:18 |
| Sandro π§ | etc-firmware.conf: β
β Ξ΅
extra-hwdb: β
β Ξ΅
extra-udev: β
β Ξ΅
firmware: β
β Ξ΅, +124.6 KiB
hwdb.bin: β
β Ξ΅, +9532.2 KiB
intel2200BGFirmware: β
β 3.1, +213.3 KiB
ipv6-privacy-extensions.rules: β
β Ξ΅
linux-firmware: β
β 20230210, +368294.5 KiB
rt5677-firmware: β
β Ξ΅, +169.9 KiB
rtl8192su: β
β unstable-2016-10-05, +274.2 KiB
rtw88-firmware: β
β unstable-2022-11-05, +298.2 KiB
udev: β
β Ξ΅, +141.1 KiB
unit-systemd-oomd.service: Ξ΅ β β
unit-systemd-udevd.service: β
β Ξ΅
wireless-regdb: β
β 2023.02.13
zd1211-firmware: β
β 1.5, +61.8 KiB
| 16:28:26 |
| Sandro π§ | thats like 350MB which we mostly don't need | 16:30:00 |
 K900 | We can just disable firmware instead | 17:28:10 |
| 15 Mar 2023 |
|  sorm_cha joined the room. | 14:07:05 |
| Sandro π§ | https://github.com/microsoft/WSL/releases/tag/1.1.5#:~:text=Do%20not%20mount%20cgroup%20v1%20for%20WSL2 | 14:19:03 |
| Sandro π§ | In reply to @k900:0upti.me
We can just disable firmware instead that is not that easy because it is all gated behind udev | 14:19:13 |
|  colamaroro joined the room. | 14:43:35 |
| 16 Mar 2023 |
| K900 | No it's not | 17:43:09 |
| K900 | hardware.firmware = mkForce [] | 17:43:13 |
| K900 | (sorry completely missed the message) | 17:43:25 |
| 17 Mar 2023 |
| Sandro π§ | I don't want to do that because undoing it is way harder than enabling udev | 01:52:21 |
| 18 Mar 2023 |
|  greg set their display name to greg. | 18:36:11 |
| 21 Mar 2023 |
| Sandro π§ | What to do when dig domain A returns not only the IP of the A record but also the A records of the ADDITIONAL ANSWER section? | 13:01:53 |
| Sandro π§ | I have some β¨nice dns serversβ¨ which think that this is necessary | 13:02:16 |
| Sandro π§ | * What to do when dig domain A returns not only the IP of the A record but also the A records of the ADDITIONAL ANSWER section and every program treats it as valid round robin addresses? | 13:03:16 |
 Cynthia | Checking the manpage, you might be able to use dig <domain> A +noadditional | 13:30:45 |
| Cynthia | I do not, however, see a way in the manpage to set it for systemd-resolved, which is probably what you'd have to set to get other programs to behave... | 13:37:03 |
| Sandro π§ | https://github.com/microsoft/WSL/issues/5806 probably better describes it | 15:53:55 |
| Sandro π§ | it mashes the two sections together to one | 15:54:05 |
| Cynthia | Wow, Microsoft baked a DNS server into the network adapter? TIL... and that's horrifying | 15:57:40 |
| Sandro π§ | Something HyperV and sometimes it falls over | 16:06:35 |
| Arson Copperflame | The DNS Server is somewhere in Hyper-Vs network stack, not the driver. I believe other Hypervisors do it in a similar fashion | 16:16:01 |
| Sandro π§ | I would just use some other DNS server but I need some IPs only available locally | 16:18:08 |
| Arson Copperflame | Sandro π§: I don't know if that helps you, but I run Tailscale on Windows and have NixOS use 100.100.100.100 as the DNS so I'm not affected by that | 16:21:15 |
| Sandro π§ | I could tinker with something like that but it probably falls apart again. I had so much trouble with DNS already. | 16:22:16 |
| Arson Copperflame | Or just set it to something like 8.8.8.8 if you can live without local host names | 16:26:23 |
| Sandro π§ | thats the problem, I cannot | 16:44:05 |
| Arson Copperflame | Is the machine portable? If no, you could set it to your router's IP | 16:44:32 |
