| 26 Jan 2024 |
 infinisil | In reply to @Ericson2314:matrix.org
https://github.com/NixOS/nixpkgs/pull/279619 is this staging backport fail innocuous? No idea, I can't spot any error really, might be worth retriggering ofborg CI (with @ofborg eval) | 01:43:41 |
 John Ericson | infinisil: I think it is just a time out because based branch is staging | 01:45:44 |
| John Ericson | but then wouldn't that affect every staging backport? | 01:45:52 |
| John Ericson | (unless we made it smart to use git mege-base release-xx staging-xx) | 01:46:10 |
| John Ericson | * (unless we made it smart to use git merge-base release-xx staging-xx) | 01:46:14 |
| infinisil | In reply to @9999years:matrix.org
would someone on the team be willing to help me get this over the line then? https://github.com/NixOS/nixpkgs/pull/280592 Also that doesn't look related to nixpkgs architecture, so I don't see why it would be relevant specifically for this team/room. For requesting reviews #review-requests:nixos.org could be used, though #dev:nixos.org could be used to discuss PRs too | 01:47:14 |
| infinisil | John Ericson: Hmm I don't think the merge base would help here, I think ofborg tests on the merged commit | 01:50:28 |
| infinisil | John Ericson: I'll retrigger it, it's very weird how the main eval status check doesn't even have any details attached | 01:51:51 |
| infinisil | (previous failure for lib/tests/release.nix: https://gist.github.com/GrahamcOfBorg/5e5e13e93fd0df5a0bfd9a7987848460) | 01:52:32 |
| infinisil | In reply to @9999years:matrix.org
would someone on the team be willing to help me get this over the line then? https://github.com/NixOS/nixpkgs/pull/280592 I need to sleep now, but it's in my notifications backlog, I'll take a look when I get to it | 01:54:27 |
| infinisil changed the room name to "Nixpkgs Architecture" from "Nixpkgs Architecture Team". | 01:57:54 |
 9999years | In reply to @infinisil:matrix.org
I need to sleep now, but it's in my notifications backlog, I'll take a look when I get to it thanks, appreciated! i'll ping #review-requests:nixos.org next time | 17:19:26 |
| 27 Jan 2024 |
|  @dooy:matrix.org changed their display name from Dooygoy to stablejoy. | 13:37:40 |
|  @lehmanator:gnulinux.club removed their profile picture. | 16:53:59 |
| @lehmanator:gnulinux.club removed their display name Sam Lehman. | 16:59:42 |
| @lehmanator:gnulinux.club left the room. | 17:03:25 |
| 28 Jan 2024 |
|  ncfavier changed their profile picture. | 14:04:01 |
|  @jade_:matrix.org joined the room. | 23:53:24 |
| @jade_:matrix.org | something I've found often very difficult to work with while doing nixpkgs work that i was reminded of in a recent review by infinisil and Robert Hensing (roberth) is that stuff that's not lib and is in fact packaging related is all slammed into all-packages.nix in the flat namespace with a bunch of packages that are not special
I'm wondering if stuff like fetchers, recurseIntoAttrs, newScope, pkgs.path and so on should get a new secondary namespace to make them easier to find out about.
also as an example, stuff like nix-update isn't obvious to new contributors who don't have long standing dms with raito to be told about them :p | 23:57:26 |
| 29 Jan 2024 |
| infinisil | jade_ Yeah fully agreed! I think we should totally have separate namespaces for the different "type" of values | 00:05:19 |
| infinisil | Kind of like a very primitive type system haha | 00:05:32 |
| @jade_:matrix.org | maybe the way to think about it also is how an ideal reference documentation would look | 00:09:33 |
| @jade_:matrix.org | and then organize based on how you'd think about that | 00:09:44 |
| @jade_:matrix.org | like fetchers, scopes, misc, idk | 00:10:32 |
| infinisil | 💯 | 00:10:50 |
| infinisil | pkgs/by-name is one small step in that direction, because that's a place where only packages are allowed, putting e.g. functions in there isn't allowed by CI | 00:11:26 |
| @jade_:matrix.org | yes good | 00:11:36 |
| @jade_:matrix.org | oh also i found a use for the isFlake feature even if my pkgs path thing gets canned: improving the errors for stuff where we ask for env-vars for broken, insecure, or proprietary packages | 00:24:04 |
 K900 | pkgs/by-name/c-/c-periphery/package.nix | 07:53:15 |
| K900 | That's a funny edge case | 07:53:17 |
