!NhAsaYbbgmzHtXTPQJ:funklause.de

Nix NodeJS

207 Members
58 Servers

Load older messages


SenderMessageTime
13 Jun 2026
@debtquity:matrix.orgdebtquity set a profile picture.03:22:58
16 Jun 2026
@scrumplex:duckhub.ioScrumplexI am working on an alternative pnpm fetcher here: https://github.com/NixOS/nixpkgs/pull/531657 It uses mitm-cache instead of creating a snapshot of the pnpm store. While there may be edge cases, from my experience with importPpnmLock.nix, it is quite robust, as pnpm only fetches tarballs in frozen lockfile mode. This may avoid the tedious fixup work we had to implement for fetchPnpmDeps and pnpmConfigHook11:15:49
@rosssmyth:matrix.org@rosssmyth:matrix.orgVery cool! I personally think mitmCache is underutilized in Nixpkgs, so it's nice to see someone using it!13:40:30
@mattsturg:matrix.orgMatt Sturgeon One issue I was running into with the Vortex package, is they pin some git+https://...#commit URLs (e.g.). Later during their build when they do a pnpm deploy, pnpm tries to re-resolve them (even with --frozen-lockfile or --offline), and fails attempting network requests. Would the new approach handle that better? 18:07:32
@scrumplex:duckhub.ioScrumplex At the moment it's only fetching tarballs as specified in the lockfile. There must be a way to prevent it from looking up Git sources though :O 18:36:13
@mattsturg:matrix.orgMatt SturgeonI know the yarn fetcher/hooks have some hacks for re-writing git-source lockfile entries, idk if the same principle applies to pnpm18:40:26
@mattsturg:matrix.orgMatt Sturgeonhttps://github.com/NixOS/nixpkgs/pull/48904018:55:04
@mattsturg:matrix.orgMatt Sturgeon* I think I'm thinking of https://github.com/NixOS/nixpkgs/pull/48904018:55:16
21 Jun 2026
@whispers:catgirl.cloudwhispers [& it/fae] joined the room.01:02:37
22 Jun 2026
@albertlarsan68:albertlarsan.fr@albertlarsan68:albertlarsan.fr left the room.17:39:00

There are no newer messages yet.


Back to Room ListRoom Version: 6