Embedded LinuxCNC distro possible?

Hi everyone,

This is going to be a, somewhat, very technical Linux question, so bare with me for a bit...

I've been tinkering with some embedded Linux experiments using musl-libc , Busybox , Toybox , etc., building a few embedded Linux systems from scratch (I've done a couple of Linux From Scratch builds, which help a lot to understand the concepts).

I started thinking "Why not try and build an 'embedded LinuxCNC' distro??".

So the idea is to try and create the most compact Linux system possible, aimed at only running LinuxCNC.
I know all about needing to satisfy dependencies and all, but there is one question:

Does anyone know if LinuxCNC depends (either directly or not) on glibc to work?

I this would help fighting the problems LinuxCNC is facing about future-proofness as well as boosting performance in RPI control based systems.

I'm hoping to have some feedback about this, I want to know if this is something that could be worth putting time into.

Miscelaneous:
There is an unmaintained, undated comparison between C/POSIX standard library implementations in Linux that shows how musl-libc compares with other libraries, actually beating glibc in some points.

My best regards,

Simão A.

We have been using Gentoo with LCNC for quite a while. So you can do this but you will very likely be on your own for support. People involved in the main project are already pretty busy.

There is already a Gentoo 64b image for the Rpi4 that makes it easy to update if you are not used to emerge.

github.com/sakaki-/gentoo-on-rpi-64bit

I'd look at this for a starting point to build something stripped down and stable for the Rpi4.

What are the issues that linuxcnc is fighting in future proof ness ?
I’m not a programmer so this question is just one of interest.

@BeagleBrainz This is the most important one, I think, the problem that LCNC relies heavily in python2 and the fact that its EOL was 1/1/2020 ( Python2 EOL Countdown , info in python.org ) and Debian is currently removing python2 support .


@Bari thank you for that, I'll definitely look into it!
I know the main project is in a bit of a pickle, especially with the python upgrade thing....
I'm going to see if I can start helping on that!
Meanwhile I left this idea here, for further discussion.

Simão A.

I am aware of the massive changes from python 2 to 3 from the very beginning of it, but i could never find a reasonable explanation on ditching the backward compatibility. Then again, since i rarely use it, i am not the one to judge.