XZ Backdoor, Nova Driver, Linux 6.9 Features & Ubuntu 24.04 Made For An Exciting April

[phoronix]

Phoronix: XZ Backdoor, Nova Driver, Linux 6.9 Features & Ubuntu 24.04 Made For An Exciting April

April 2024 is now in the books after writing 257 original Linux/open-source-related news articles and another 13 featured articles / Linux hardware reviews. Here's a look back at the most exciting (popular) content from April...

https://www.phoronix.com/news/April-2024-Highlights

[Kjell]

Lennart Poettering is the founder of systemd and a Microsoft employee.

His vision for systemd is to encapsulate the entire ecosystem into one monolithic setup by systematically converting parent distributions to fully specced systemd machines. Biggest distributions are gradually deepening their integration (like Debian, Arch Linux, Fedora & OpenSuSE) by forcing you to use systemd toolset to build or even run, like: systemd-run, systemd-libs, libsystemd, systemd-resolvconf & systemd-sysvcompat. This is while replacing core system components with systemd equivalents: systemd-boot, systemd-logind, systemd-homed, systemd-networkd, systemd-resolved, systemd-nspawn & systemd-cryptenroll..

This gives one entity too much power and the ability to introduce compatibility breaking features. Yes, today it's still possible to opt-out of some of systemd's components if you compile manually (not to mention most distributions are bundling the entirety of systemd). However, if you've ever interacted with systemd developers in the early days you'd know how stubborn they are due to their vision. There is no negotiating with these people.

You could argue this is social engineering which is slowly steering the ecosystem in a specific direction which will eventually fragment the community into systemd/GNU/Linux vendor lock-in, that is when inevitably ecosystem specific features get introduced. Slowly enough to not cause too much disruption hence result in acceptance.

We'll see a lot more exploits akin to XZ as the bloat continues.

[Kjell]

In my opinion to stop the spread of systemd, there needs to be a set of agreed upon protocols and standards similar to the way Wayland is designed.

Common design principles would allow systemd components to be (forever) interchangeable without fragmenting the community into systemd/GNU/Linux.