| 6 Dec 2024 |
 Ramses 🇵🇸 | I just built it from main like 2h ago or such | 09:46:11 |
 aloisw | In reply to @rvdp:infosec.exchange
It seems that something changed in the handling of builtins.currentSystem in lix? Doing something like nix build -f . wstunnel in a nixpkgs checkout now fails with error: attribute 'currentSystem' missing, while it worked before and also works correctly with cpp nix Can't reproduce, but that error smells of pure evaluation getting enabled somehow. | 10:47:19 |
| Ramses 🇵🇸 | Hmm, that's pretty weird | 11:02:52 |
| Ramses 🇵🇸 | I thought -f explicitly disables pure eval mode | 11:03:35 |
| aloisw | In reply to @rvdp:infosec.exchange
I thought -f explicitly disables pure eval mode This appears to be correct. So even if the setting got applied somehow (e.g. config file), you should not see this error. | 11:07:33 |
 antifuchs | sayyyy, what "type" values are in a flake lock file? Is there a list in source code somewhere? I see "github" and "path" ones, but there must be others, right? | 13:18:12 |
| antifuchs | * sayyyy, what "type" values are in a flake lock file? Is there a list in source code somewhere? I see "github", "tarball" and "path" ones, but there must be others, right? | 13:19:18 |
 @acidbong:envs.net | In reply to @antifuchs:asf.computer
sayyyy, what "type" values are in a flake lock file? Is there a list in source code somewhere? I see "github", "tarball" and "path" ones, but there must be others, right? different URLs, for tarballs, local paths (git or not) or Github/Gitlab/Sourcehut repos | 13:30:03 |
| @acidbong:envs.net | sadly, there's no Gitea/Forgejo scheme yet | 13:30:22 |
| @acidbong:envs.net | * sadly, there's no Gitea/Forgejo schema yet | 13:30:31 |
| antifuchs | I see some indirect things in the libfetchers/ source tree | 13:30:38 |
| antifuchs | sounds like "path" is the only thing I need to really guard against here | 13:30:55 |
|  Jana changed their profile picture. | 14:30:13 |
| Jana changed their display name from jana to Jana. | 14:30:16 |
|  Nico 🐉🔥 joined the room. | 16:05:40 |
 alina🏳️⚧️🐾 | i only now realized lix is in nixpkgs now and i don't need to re compile it on every update anymore awa | 16:41:59 |
 famfo | hi, is cache.lix.systems down? I cannot reach it anymore (v4 and v6) | 19:46:24 |
| famfo | 
Download clipboard.png | 19:46:29 |
 Grimmauld | Uh, i somehow can't define recursive functions in lix repl, is this intended? | 22:26:18 |
 raitobezarius (DECT: 3538 / EPVPN 2681) | In reply to @famfo:ccc.ac
hi, is cache.lix.systems down? I cannot reach it anymore (v4 and v6) Let me check | 22:28:06 |
| raitobezarius (DECT: 3538 / EPVPN 2681) | In reply to @famfo:ccc.ac
hi, is cache.lix.systems down? I cannot reach it anymore (v4 and v6) It's up on my side | 22:28:41 |
 szlend | You can work around it at least: fac = let fac = n: if n == 1 then 1 else n * fac(n - 1); in fac | 22:33:27 |
 aftix | isn't that just how the nix language works? | 22:34:00 |
| szlend | Behaviour seems the same in Nix, so not a regression or anything | 22:34:02 |
| szlend | more like nix repl | 22:34:28 |
| szlend | x = 1 in repl is not the same as a let binding | 22:34:48 |
| famfo | In reply to @raitobezarius:matrix.org
It's up on my side Hm, Scaleway doesn't like NetCologne, still unreachable from my side | 22:44:09 |
| famfo | Reachable from Hetzner though, interesting | 22:44:31 |
 puck | i'd guess it's probably an internal scaleway routing issue? | 22:45:30 |
| famfo | either that or the reverse route somewhere | 22:46:15 |
