| 22 May 2024 |
 x10an14 | * Hey y'all!
I'm working on nixifying a Python application, and chose to use poetry2nix due to previous familiarity.
Originally the project used hatch (not really relevant).
However, the end-result is a docker image (no surprises there).
And the docker image previously produces has much smaller size than the one generated by Nix.
(I'm leveraging nixpkgs's dockerTools.buildImage on the mkPoetryApplication).
We're talking like 125MiB vs 527MiB.
| 11:30:10 |
 K900 | Probably optional dependencies somewhere | 11:30:37 |
| x10an14 | Some of it might be explained by the python dependencies in the nix docker image not being compressed (as they seem to be in the original one).
But what gives? Any options? | 11:30:43 |
| x10an14 | In reply to @k900:0upti.me
Probably optional dependencies somewhere I made sure to specify the groups = [] list in mkPoetryApplication, so that can't be it, right? Or am I mistaken? | 11:31:10 |
| K900 | Optional dependencies of dependencies | 11:31:21 |
| K900 | Possibly | 11:31:23 |
| K900 | But it's hard to tell | 11:31:25 |
| K900 | nix-tree? | 11:31:26 |
| x10an14 | Need to read up on nix tree, I thought i could just give it my results folder, but alas | 11:32:56 |
| x10an14 | 
Download 2024-05-22T13:44:20,258284425+02:00.png | 11:45:01 |
| x10an14 | According to nix-tree, I live in a world of impossibilities 🙃 | 11:45:02 |
| x10an14 | Does this make any sense to you? | 11:45:10 |
| K900 | You're looking at the drv file I think | 11:47:39 |
| x10an14 | Hmm, okay. I guess a foray into learning nix-tree is next then, if I can afford the time. | 11:48:40 |
|  NixOS Moderation Botchanged room power levels. | 15:25:57 |
