| 19 Dec 2025 |
 whispers (it/fae) | huh, i stand corrected | 18:35:38 |
 Rutile (Commentator2.0) feel free to ping | where can i send money to support lix / -infra? | 19:04:19 |
 neobrain | Heh I see, yeah I can relate how it's not the most pressing task to sit down and write automation for π
| 19:36:29 |
|  PopeRigby changed their display name from PopeRigby to "PopeRigby (UTC-8)". | 20:01:58 |
| PopeRigby changed their display name from "PopeRigby (UTC-8)" to PopeRigby. | 20:03:08 |
|  Bakbukdibbuk joined the room. | 21:27:17 |
| 20 Dec 2025 |
 just1602 | In reply to @k900:0upti.me
OK that's weird and I guess I missed this somehow I think the idea was to have 2.92 and then 2.93 as latest but keep 2.91 as stable due to the regression present in latest. | 05:56:13 |
|  Acid Bong joined the room. | 07:20:47 |
| Acid Bong | moin. i noticed that Nix and Lix do nix flake check differently: Nix evaluates everything in alphabetic order, but Lix's order is different. what's the specifics of this? | 07:26:02 |
 Sofie π³οΈββ§οΈ (she/her) | Me too is wondering this | 09:20:48 |
| Sofie π³οΈββ§οΈ (she/her) | I have hardware too ;) | 09:21:01 |
 K900 | Hardware is hard because of trust concerns | 09:49:28 |
| K900 | For money there's afnix opencollective | 09:49:50 |
| K900 | https://opencollective.com/afnix/donate | 09:49:50 |
| K900 | But I'm not sure there's any particular expenses that will go to right now, you'd have to ask probably @raitobezarius | 09:50:17 |
 raitobezarius | I'd argue that money and hardware is not something we are lacking of | 11:08:36 |
| raitobezarius | Hands on the project or joining the AFNix infrastructure team as a volunteer is probably something we would appreciate even more | 11:09:03 |
| Sofie π³οΈββ§οΈ (she/her) | Agree | 11:18:00 |
| Sofie π³οΈββ§οΈ (she/her) | I would be looking to join the infra team, just as a volunteer, for now | 11:18:14 |
| Sofie π³οΈββ§οΈ (she/her) | Is there ANYTHING spare hardware can be used for? | 11:18:34 |
| Sofie π³οΈββ§οΈ (she/her) | Just as a general question | 11:18:56 |
 0x4fbb09 it/its β―βΞΞ | untrusted spare hardware would be useful for running tests on, no?
fuzzing / reproducibility checks / maybe PR CI
unsure how many of those lix has a need for but none of those are sensitive, the worst a hostile machine could do is lie and say "yeah there's (no) bugs" | 11:41:37 |
 Lotte (it/its)/Cinny (she/her) | i donβt think nix lets you split tests in that way without going full on cppnix packaging | 11:43:25 |
 tea | I wonder, has anyone made a nix pinning system that vendors code, node-modules style? | 11:46:54 |
| tea | as far as I know, things like npins generate a nix file with fetchers; it'd be an interesting alternative approach | 11:47:52 |
| 0x4fbb09 it/its β―βΞΞ | "npins but the fetchers are just a path" ig? | 11:48:24 |
| tea | well yeah, but I'd also want it to auto-vendor the deps :)
not sure this would be a practical improvement over npins, but maybe its good to avoid ifd | 11:49:47 |
| Lotte (it/its)/Cinny (she/her) | crate2nix does that for some reason | 11:50:50 |
| 0x4fbb09 it/its β―βΞΞ | i don't think npins strictly speaking uses ifd, it just uses builtins fetchers? | 11:52:08 |
| 0x4fbb09 it/its β―βΞΞ | (which you can get it to not do, if you give it a nixpkgs, but you need to use a builtin fetcher to fetch nixpkgs itself) | 11:52:34 |
