boostrap VS mach-nix

I've got something very similar! https://github.com/toasterrepairman/boostrap/tree/main/nix

It's also interesting to hear comments from the Nix detractors here. There's ample room to criticize the language and even the design philosophy behind the package manager, but it's a shame that most people never give it a "proper shot". Maybe that's the fault of Nix maintainers though.

Nvidia works excellent in my experience, Nix provides the best Wayland/Nvidia experience I've had out of any distro. I'm using a 3070ti, but you should be able to steal most of my configuration if you want a similar setup: https://github.com/toasterrepairman/boostrap/blob/main/nix/n...

Another way if the above doesn't work is to use mach-nix:https://github.com/DavHau/mach-nix

For python, I can highly recommend mach-nix. Makes it really easy to also keep a requirements.txt to stay compatible with non-nix-users.

The ESP32 toolchain is quite cumbersome to install - even under NixOS. The wiki tells you to create a shell.nix that creates a python virtual env at the first execution. I used DavHau's mach-nix to create esp-idf.nix and subsequently micropython-esp32.nix.

mach-nix and friends are pretty good, but I've found that the expectations of all the commonly-used data science stuff is pretty antithetical to the Nix Way. I'm not sure if it's still the case, but last I tried, Hydra wasn't building CUDA (since it's non-free), so I had to compile that and e.g. TensorFlow and PyTorch. Very painful, even on a beefy system.

I don't do much Python, but usually everything works just fine with mach-nix. I would try something like this: ``` { pkgs ? import (fetchTarball "https://github.com/nixos/nixpkgs/archive/nixpkgs-unstable.tar.gz") { } }:

There are some sifferent/new tools for adding your own Python packages these days. It's still not truly solved, but one of these new package generation tools might serve your better:

poetry2nix: https://github.com/nix-community/poetry2nix

dream2nix: https://nix-community.github.io/dream2nix/guides/getting-sta...

mach-nix: https://github.com/DavHau/mach-nix

pip2nix: https://github.com/nix-community/pip2nix

pynixify: https://github.com/cript0nauta/pynixify

The tools available to you at the time (pypi2nix and maybe python2nix, if it was a long time ago) have been abandoned in favor of the newer tools, I think chiefly poetry2nix but I'm not sure.

There's still the Nixpkgs buildPythonPackage stuff, I think, if your goal is to upstream a lib into Nixpkgs. But if you just want to build your own Python applications and vendorize the deps (e.g., for work), you might try one of the tools above, which weren't available 3+ years ago. Maybe Nixy Python users and developers can reply with some of their experiences using those tools :)

Have a look at mach-nix which is a small utility library for nix to build Python packages declaratively.

final: prev: { xonsh = prev.xonsh.overrideAttrs (old: rec { version = "0.12.4"; src = final.fetchFromGitHub { owner = "xonsh"; repo = "xonsh"; rev = version; sha256 = "0kdps0gf0767zy0fs6qn39rv4z3x7ck0qz1pzx6962593171yk8b"; }; propagatedBuildInputs = prev.xonsh.propagatedBuildInputs ++ [final.python3Packages.virtualenv]; }); python39 = prev.python39.override { self = prev.python39; packageOverrides = python_final: python_prev: { prompt-toolkit = python_prev.prompt-toolkit.overrideAttrs (old: rec { version = "3.0.29"; src = final.python3Packages.fetchPypi { pname = "prompt_toolkit"; inherit version; sha256 = "sha256-vWQPYOjOzXTw3CSXE9QzrOLdxitl7gf5bTWOCxUrbqc="; }; }); }; }; # Using mach-nix to fetch unpackaged xontrib plugins # adapted from https://github.com/NixOS/nixpkgs/issues/75786#issuecomment-873654103 mach-nix = import (builtins.fetchGit { url = "https://github.com/DavHau/mach-nix/"; ref = "refs/tags/3.4.0"; }) { pkgs = final; }; xonsh_pyenv = final.mach-nix.mkPython { requirements = '' xontrib-fzf-widgets xonsh-direnv ''; }; xonsh_with_plugins = final.xonsh.overrideAttrs (old: { propagatedBuildInputs = old.propagatedBuildInputs ++ final.xonsh_pyenv.python.pkgs.selectPkgs final.xonsh_pyenv.python.pkgs; }); }