guix-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Improving importers best investment for growing gnu/packages/


From: zimoun
Subject: Re: Improving importers best investment for growing gnu/packages/
Date: Mon, 02 May 2022 11:19:15 +0200

Hi,

On Tue, 29 Mar 2022 at 10:50, Brendan Tildesley <mail@brendan.scot> wrote:

> Therefore I'm working on making use of
> https://github.com/rust-lang/crates.io-index  to fully import base definitions
> all required crates rather than pulling metadata from the internet with each 
> refresh.

>From my point of view, the issue is that it requires external Rust tools
and I think such tool will never be part of the Guix dependencies.  Guix
itself cannot depends on Rust-thing, GHC-thing, OCaml-thing,
Julia-thing, etc. because they ease the importers.

And that’s why the current importers parse metadata and somehow
“reimplement” the logic behind the package manager of each language.  I
agree that it is a lot of work and hard to maintain but doing so allow
Guix to have a small footprint (dependency, speaking).

That’s said, how to improve the importers?

Well, Guix provides now GUIX_EXTENSIONS_PATH which allows to extend by
other subcommands.  “guix workflow” is an example, I do not remember how
“guix home” did initially, anyway!

My point is: this mechanism allows to have a package (with many
dependencies) extending Guix itself.  Somehow, it reads

    guix install guix-rusty
    guix rusty <stuff>

where ’guix-rusty’ would be a Guile package (program) using whatever it
needs from the Rust ecosystem (say parse system calls of cargo or
whatever).

Guix is Scheme library after all, let use it! :-)


Cheers,
simon



reply via email to

[Prev in Thread] Current Thread [Next in Thread]