Is Nickel intended to eventually be a Nix (lang) alternative/replacement?

As the descriptions say, Nickel is still very “experimental”, however its genetic “close-ness” to Nix is clear. Part of me wonders, that if all goes perfectly well, if Nickel will only live in the realm of responsibility of config super-langages like Dhall, or if there is possibly a desire of Tweag-and-company to have it interop with or succeed Nix-the-languge in some way?


Gleaning from Figure out an adoption plan · Issue #93 · tweag/nickel · GitHub

It seems the plan very well may be to succeed Nix at some point.


I’m curious if it’s ready to use in our nix configs, if there’s already examples and how nice it looks.


Due to lack of documentation in the current nix ecosystem, a strongly typed language would save my ass


I think Nickel is intended to be an attractive choice for the Nix community to use in conjunction with the rest of the Nix ecosystem, including potentially to generate or replace code written in the Nix language.

It’s probably not worth playing with for use in NixOS-like module systems until the override semantics are settled.

Once it does have merge support, it might be fun to see what it feels like to write code that produces Nix derivations using NixOS-like merges instead of packages-as-functions like in Nixpkgs, or to see if it can be used to create a drop-in replacement for NixOS modules.

If Nickel can produce better error messages than Nix does and provide additional provenance information for the output of NixOS-like merges, it really would be attractive option among languages to use with the Nix ecosystem.

Hosted by Flying Circus.