| 8 Mar 2024 |
 @jade_:matrix.org | yeah | 21:15:59 |
 infinisil | programs.sqlite is an example, and it's really valuable | 21:16:04 |
| infinisil | Here's a smallish project: Replace programs.sqlite with nix-index | 21:16:29 |
| infinisil | Would deprecate https://github.com/Mic92/nix-index-database and give everybody a nix-index database by default | 21:16:50 |
| infinisil | Well, as long as you fetch channels and not git branches | 21:17:00 |
| @jade_:matrix.org | hm. channels have been killed, but we could make a hydra job for the nix-index of the current channel version.... or something? | 21:17:11 |
| @jade_:matrix.org | well | 21:17:14 |
| @jade_:matrix.org | flakes have killed channels | 21:17:18 |
| @jade_:matrix.org | whether that was actually advisable or not is another question hehehe | 21:17:25 |
| infinisil | They only killed nix-channel! | 21:17:28 |
| infinisil | Channels are still very much needed and alive | 21:17:36 |
| infinisil | Well not that much actually | 21:18:04 |
| @jade_:matrix.org | (nix-channel and the idea of mutable channels is like. genuinely Bad, but it would be nice if we could have postprocessed versions of nixpkgs instead of just source code, that are actually pinned) | 21:18:11 |
| infinisil | At least programs.sqlite clearly demonstrates that channels can have huge value. And flakes could make use of that too | 21:18:33 |
| infinisil | I like how npins supports channels natively | 21:19:44 |
| infinisil | So if you do npins add channel nixpkgs-unstable and use the resulting Nixpkgs to build your NixOS system, you get an prebuilt always-up-to-date binary lookup for free! | 21:20:20 |
| infinisil | Without nix-channel! | 21:20:43 |
|  ma27 joined the room. | 21:47:40 |
 tomberek | Apologies for the re-start of a topic, but regarding the mentions above for makeScope, I've experimented with a variant of it that can be used with nested attrsets that was quite pleasant to use (ie; the // was not destructive and allowed for adding things to pythonPackages in a composable manner). +1 on the topic of improving the usage of package sets. | 22:48:13 |
 Growpotkin | Please - let me split `makeScope` into reusable and well documented helpers.
| 22:50:15 |
| tomberek | and use them in Nixpkgs, to help make everything behave consistently? (asking for a friend ;) ) | 22:51:29 |
| Growpotkin | That function is THE CORE of `nixpkgs` and currently it's really really hard to grok.
It legitimately took me like 6 months of fiddling with similar routines to realize "I just spent months rewriting `makeScope`, and now I actually understand what it was doing" 🤣 | 22:52:04 |
| Growpotkin | It's a wicked useful routine and it's subparts are also incredibly useful in isolation.
In nixpkgs the parts are always used as a single long pipe, but dream2nix, floco, and hella other `nix` frameworks have needed to rewrite subparts of `makeScope` because it's one pipe | 22:54:21 |
| 9 Mar 2024 |
| @jade_:matrix.org | i had to stare it down to make a subsidiary scope of a customized pythonPackages. very wacky function, infinite appreciation to Valentin for getting a docs pr happening for it | 00:22:29 |
 raitobezarius | maybe of note: https://github.com/NixOS/nixpkgs/pull/87388 | 00:23:35 |
| raitobezarius | and I realize I missed to answer to Valentin's message here 🤦♂️ | 00:23:43 |
|  Yinfeng joined the room. | 02:27:52 |
|  Benedikt joined the room. | 15:38:13 |
| 10 Mar 2024 |
 fgaz | FYI I'm working on a phorge package and module, hopefully I'll publish it soon | 06:46:24 |
 adisbladis | You can't do that with Python specifically because of collisions | 07:00:34 |
