| 20 May 2023 |
 6t8k | * Ah, sorry, I thought it would be OK to ask this here because the Home Manager repo is under the nix-community GitHub org and didn't know there is a more specific room. Will ask there, thanks. | 19:54:39 |
 aidalgol | In reply to@6t8k:matrix.org
Ah, sorry, I thought it would be OK to ask this here because the Home Manager repo is under the nix-community GitHub org and didn't know there is a more specific room. Will ask there, thanks. AIUI, this channel is really just for meta discussion about the nix-community org itself. | 20:32:20 |
| 21 May 2023 |
 DavHau | Maybe some admin can add this information to the channel description. | 08:29:31 |
 adisbladis | I could do it, but how would it be phrased? | 09:09:27 |
| DavHau | Maybe:
Meta discussions about the nix-community org. (For project specific discussions use github issues or projects own matrix channel) | 09:46:18 |
| adisbladis changed the room topic to "Meta discussions related to https://nix-community.org. (For project specific discussions use github issues or projects own matrix channel)" from "Discussions related to https://nix-community.org/ ". | 11:17:20 |
| adisbladis | In reply to @hsngrmpf:matrix.org
Maybe:
Meta discussions about the nix-community org. (For project specific discussions use github issues or projects own matrix channel) I made a slight variation just to get a clickable link, thanks! | 11:18:04 |
| 22 May 2023 |
|  lunik1 joined the room. | 12:47:11 |
| 23 May 2023 |
|  @daedmon:matrix.org joined the room. | 21:21:49 |
| 25 May 2023 |
|  raitobezarius changed their display name from raitobezarius to disko in NixOS 23.11 when. | 13:32:19 |
| raitobezarius changed their display name from disko in NixOS 23.11 when to raitobezarius. | 13:37:26 |
| 26 May 2023 |
|  @lotte:chir.rs changed their profile picture. | 18:41:00 |
| 27 May 2023 |
|  NixOS Moderation Botchanged room power levels. | 16:40:24 |
| 28 May 2023 |
 @fedx:matrix.org | Messing with the NUR and I cannot figure out how to take one package as an input of another within the same repo. | 16:15:53 |
| @fedx:matrix.org | More specificaly, I need to create a builder to simplify a standard build process, and cannot figure out how to get that into another package. | 16:22:02 |
| @fedx:matrix.org | Wait, would I use something like mypkg = pkgs.callPackage ./path/to/thing {inherit myBuilder};? | 16:40:52 |
| 29 May 2023 |
|  Mic92 invited  Yingchi Long. | 09:23:34 |
| Yingchi Long joined the room. | 09:59:52 |
| Yingchi Long | Hello guys! I would like to migrate my recently wrote language server (WIP) into nix-community incubator https://github.com/inclyc/nixd . I've sent emails to administrators, and what should I do now?🙂 | 10:52:23 |
| 30 May 2023 |
 daxvena | oh i'm not sure, but wow this looks amazing Yingchi Long!! | 00:19:10 |
| daxvena | i'm guessing the support for cross-file analysis doesn't do dataflow analysis to infer function argument types from usage? if it did that would be magical, but even if not, this project looks really cool! i'll definitely have to check it out! | 00:19:33 |
| daxvena | * i'm guessing the support for cross-file analysis doesn't do dataflow analysis to infer function argument types from usage? if it did that would be magical, but even with that, this project looks really cool! i'll definitely have to check it out! | 00:19:58 |
| daxvena | * i'm guessing the support for cross-file analysis doesn't do dataflow analysis to infer function argument types from usage? if it did that would be magical, but even without that, this project looks really cool! i'll definitely have to check it out! | 00:20:10 |
| Yingchi Long | In reply to @daxvena:matrix.org
i'm guessing the support for cross-file analysis doesn't do dataflow analysis to infer function argument types from usage? if it did that would be magical, but even without that, this project looks really cool! i'll definitely have to check it out! Well, this basic idea is, nixd allow you configure a global installable which has exactly same argument as nix eval. For each evaluation, we inject nix parsing & eval cache with our own AST structure that tracing eval information. If the file is a lambda (e.g. nixos modules), it will be evaluated with real arguments, and we can know which argument is passed to it. Generally lambdas could be invoked many times, nixd stores the lastest information. | 01:42:49 |
| daxvena | oh ok!! so it doesn't merge each possible call, but that sounds way better than nothing!! that sounds amazing! | 01:43:58 |
| Yingchi Long | 
Download Screenshot_20230528_124426_20230530094521.png | 01:45:39 |
| Yingchi Long | 
Download Screenshot_20230528_124528_20230530094524.png | 01:45:44 |
| daxvena | that's really cool! i've been wanting this so badly 😅 | 01:46:43 |
| Yingchi Long | In reply to @daxvena:matrix.org
that's really cool! i've been wanting this so badly 😅 For example you can inspect real arguments for callPackage & nixosModule? | 01:47:54 |
| Yingchi Long | AST injections now is generally pure magics. We have our own classes inherited from nix itself, and override its `eval` method.
https://github.com/inclyc/nixd/blob/2ee1ef796dd6dd5f0bc8cc4a3bc804972018c973/lib/nixd/include/nixd/CallbackExpr.h#L21
https://github.com/inclyc/nixd/blob/2ee1ef796dd6dd5f0bc8cc4a3bc804972018c973/lib/nixd/src/CallbackExpr.cpp#L41 | 01:50:43 |
