| 13 Oct 2025 |
 raitobezarius | the heuristic is completely based on which resolver address family you have | 18:43:07 |
| raitobezarius | so if you use resolved and you are IPv6-only, you will end with the wrong result | 18:43:16 |
| raitobezarius | (and vice versa) | 18:43:18 |
| raitobezarius | But that's not the only problem we are seeing I fear here | 18:43:28 |
| raitobezarius | I surmise that rewrites are failing and that the /etc/resolv.conf is completely wrong | 18:43:45 |
| raitobezarius | do you have bash or something? | 18:43:52 |
| raitobezarius | i mean | 18:43:55 |
| raitobezarius | you can always do that once you have a bootstrap | 18:44:00 |
 Marie | yes | 18:44:01 |
| Marie | i have working cross | 18:44:05 |
| raitobezarius | but I guess it's the powerpc machine | 18:44:07 |
| raitobezarius | yeah but working cross won't tell us enough haha | 18:44:18 |
| raitobezarius | i mean | 18:44:23 |
| raitobezarius | you can always run cross bash as a builder and report the contents of that file | 18:44:30 |
| raitobezarius | this should be a release blocker because everyone is running into it | 18:45:01 |
| raitobezarius | https://git.lix.systems/lix-project/lix/issues/1014 | 18:47:55 |
| Marie | eval is so incredibly slow on this machine, it's insane | 18:55:39 |
| Marie | nix-build --expr 'let pkgs = import ./nixpkgs.nix; in derivation { name = "resolvconf"; builder = pkgs.lib.getExe pkgs.bash; args = [ "-c" "${pkgs.coreutils}/bin/cat /etc/resolv.conf > $out" ]; system = "powerpc-linux"; outputHashAlgo = "sha256"; outputHashMode = "flat"; outputHash = "sha256-699WAnKnc1cZXDnpg0C3fhjIqM4gJe6VDp4MewFGerg="; }'
I think this should be correct?
results in | 19:04:23 |
| Marie | nameserver 127.0.0.53
options edns0 trust-ad
search .
| 19:04:36 |
| Marie | and the usual resolved comment | 19:04:49 |
 aloisw | In reply to @raitobezarius:matrix.org
this should be a release blocker because everyone is running into it Release blocker for what? 2.93.3 already has pasta. | 19:07:24 |
| raitobezarius | for 2.94.0 | 19:07:34 |
| raitobezarius | 2.93.3 doesn't enable pasta outside of unstable, no? | 19:07:44 |
| raitobezarius | (ideally, we should fix it for 2.93.4 as well) | 19:07:51 |
| aloisw | In reply to @raitobezarius:matrix.org
2.93.3 doesn't enable pasta outside of unstable, no? No, the in-tree packaging always enables pasta on Linux. | 19:09:56 |
| raitobezarius | including on 25.05? | 19:10:07 |
| raitobezarius | spicy, if so | 19:10:21 |
| aloisw | It ships its own pasta. | 19:10:32 |
| raitobezarius | doesn't change much to what we should do anyway | 19:10:32 |
| raitobezarius | i have it on my list to squash the final usability bugs of pasta | 19:10:52 |
