| 6 Apr 2023 |
 j-k | the script takes fooooreeeeeveeeeer to run.
I think nixpkgs-updates avoids any of the big language attributesets like JS & haskell AFAIK | 14:49:16 |
 hexa | it works for python packages, which is kinda big | 14:53:13 |
| hexa | but if you care about a package within nodePackages, and it uses packages-lock.json (aka npm), convert it to use buildNpmPackage and use nix-update as the updateScript | 14:53:56 |
| j-k | well for python most things are standalone right? like the package has it's own src. rather than a mega sources.json that gets mapped out to many packages? | 14:54:31 |
| hexa | yep | 14:55:33 |
 Lordie | In reply to @hexa:lossy.network
but if you care about a package within nodePackages, and it uses packages-lock.json (aka npm), convert it to use buildNpmPackage and use nix-update as the updateScript do you mean, pulling it out of nixpkgs/pkgs/development/node-packages into nixpkgs/pkgs/development/foo/bar and make a derivation using buildNpmPackage instead? | 15:02:01 |
| Lordie | In reply to @hexa:lossy.network
but if you care about a package within nodePackages, and it uses packages-lock.json (aka npm), convert it to use buildNpmPackage and use nix-update as the updateScript * do you mean, to pull it out of nixpkgs/pkgs/development/node-packages into nixpkgs/pkgs/development/foo/bar and make a derivation using buildNpmPackage instead? | 15:02:25 |
| hexa | yes | 15:02:42 |
| hexa | nodePackages is a disaster | 15:02:51 |
| Lordie | huh so for non-lib node packages, nixpkgs/pkgs/development/node-packages is sort a legacy way of doing things | 15:04:20 |
| Lordie | I'll make a PR, thanks | 15:04:59 |
| hexa | it made itself legacy due to the time required for generate.sh to run | 15:09:49 |
| hexa | also sometimes it just fails 2 hours in | 15:09:55 |
| hexa | the rough idea was to share package dependencies | 15:10:12 |
 Sandro 🐧 | node2nix does no parallelism | 15:38:25 |
| 12 Apr 2023 |
|  Yuddite G joined the room. | 09:11:02 |
| 16 Apr 2023 |
| Yuddite G changed their profile picture. | 23:09:30 |
| 20 Apr 2023 |
|  @federicodschonborn:matrix.org changed their display name from Federico Damián Schonborn to Federico Schonborn. | 01:02:28 |
| 25 Apr 2023 |
|  @trofi:matrix.org left the room. | 15:45:11 |
| 26 Apr 2023 |
| Yuddite G changed their display name from Yuddite Pilot to Yuddite Groyper. | 04:49:01 |
| Yuddite G changed their display name from Yuddite Groyper to Yuddite G. | 21:02:19 |
| 28 Apr 2023 |
|  jopejoe1 joined the room. | 12:16:14 |
| 1 May 2023 |
 vcunat | Not a problem, just got surprised now by the choice of attribute alias
arcanPackages.harfbuzz: 7.1.0 -> 7.2.0 #228746
| 08:05:37 |
 ryantm | Yeah, there is an idea to try to find the shortest atterpath for a given derivation, but we haven't done it. | 12:04:36 |
 Artturin | better just fix that in nixpkgs | 17:00:28 |
| 4 May 2023 |
|  raitobezarius joined the room. | 20:45:15 |
|  Ambroisie joined the room. | 20:46:14 |
| Ambroisie | Reposting here from #dev:nixos.org
<@ambroisie:belanyi.fr> That'd be cool, on an opt-in basis
(If you mean r-ryantm notifying when update is available but PR could not be opened, yes that would be quite lovely)
Is that something that could be done easily? (never looked at the code of the bot) | 20:52:59 |
|  maralorn joined the room. | 20:56:29 |
| Ambroisie | If so, I'll open an issue about it (my yak shaving is currently aimed at other goals unfortunately ľ | 20:57:31 |
