Its yet another reason why I recommend Nix or Guix to more tech-savvy Linux users. It probably has some of the most intuitive packaging UX I've ever used, so even a non-technical user should be able to package their own stuff given time enough to read the man pages and use working examples, or via v*be coding. It is probably the reason for why there are so many third party package channels. There's even a webring through which you can easily look most of them up, provided they've submitted a pr to the keyring: https://toys.whereis.social/
The niggercuck clowns behind Flatpak are planning to make a new version which will be coded in Rust cause "Gamers and Rust Nerds will save Linux", will require Wayland and Systemd, and will remove x11 support because "no one uses it".
I really don't understand this criticism of Devuan. If someone wants to run packages that have only been tested about ten time '''as much as the core Arch system, they can just start running unstable. Personally, I don't see any reason you'd need the very latest version of i3-wm, but, unstable had builds of that within a week of release.
Except for drivers there's not much that actually changes that matters. Web browsers you can use their repo. Other stuff that you want closer to bleeding edge there are appimages or containers for server stuff. But the OS can just stay stable.
devuan continues winning
no systemd, no age verification, no rust and everything works just fine like a normal debian install
just enjoy your five year old packages
i haven't had any issues with dependencies on systemd, i think the last problem i had was with getting xpad drivers to work with dolphin emulator but that was years ago and i didn't really care that much to being with.
we might not have to "fight" it if we can fake it. The biggest and hardest thing to possibly emulate would be all the IPC/network socket/cgroup stuff. The whole dependency graph can be theoretically handled by autistic shell script control where you manually control the order each service starts.
I mean... systemd already won so it's not like our objections will ever change that. Even so, people still hold onto the objections precisely because long-telegraphed issues that people brought up 12-15 years ago have come up time and again.
In context, back in 2011-2012 when systemd first entered the scene and began causing a huge uproar, it was one option among many alternatives to sysvinit. This was during a time when RHEL6 shipped with Canonical's upstart and Gentoo's OpenRC was finally starting to evolve into an init system instead of being a layer atop sysvinit. The problems before systemd were obvious: we needed something hotplug-capable, consolekit was doing tons of stuff outside its intended scope, all the workarounds were becoming untenable to maintain long-term, and something needed to be done.
If systemd had stayed constrained purely to being simultaneously an init and service manager, and never evolved past that original scope... it would actually be pretty damn good. dinit shows us what a constrained systemd-like init framework could've looked like, but it arrived far too late. systemd's also legitimately worth getting pissed off at considering how much vertical and horizontal integration happened over the years. The sheer scope of systemd means that a bug in one obscure section that only handles $insert_function_here can, and often does, ripple into other wholly unrelated sections. Do these bugs and even the CVEs get patched relatively quickly? Yes. Were these bugs 100% avoidable if systemd remained constrained to a much more narrow and focused scope? Also yes.
Any retard can tell you that fighting against systemd is futile. We all know it's futile, and the only people delusional enough to think otherwise are either way too deep in the Gentoo/Artix rabbit holes, abandoned ship to the BSD camp a while ago, or they're just flat-out trolling you to get a rise.
Only issue I've ever had with software running on non-systemd distros is with one program that was forced to use ALSA instead of PulseAudio.
Long story short, I wasn't able to record audio coming from my computer with Audacity (or even Sneedacity) with the program that was forced on ALSA so I had to get creative and take a double-male audio jack to connect to the "Speaker" and "Microphone"/"Line-In" ports (I forget which one it was) on the back of the computer and record the audio that way.
For me, the biggest issue was getting Flatpak'd software to run at all. I had to make some modifications to /etc/security/limits.conf or whatever the file was when I used Artix for a time. That was what ultimately convinced me to move over to AppImages over Flatpaks, and given how Flatpaks are now getting more intimately married to systemd, that was arguably the correct call to make.
i haven't had any issues with dependencies on systemd, i think the last problem i had was with getting xpad drivers to work with dolphin emulator but that was years ago and i didn't really care that much to being with.
Lots of minor issues with third-party packages that assume that you are running it and have post install scripts that invoke it.
I don't mind fixing my own problems that I created on my computer, but introducing more problems unnecessarily for myself is not something I am willing to do.
I will probably watch the video later after work. However I noticed he put Debian as "meh", I looked at his complaints. His complaint about a number of languages/compilers being behind the latest while technically correct is misleading. It is trivial to install these outside apt and is often the recommended way. It is a total non-issue.
Their site recommends using rust-up and I know that this works correctly as I used it when learning the language a year ago. It's the same for many other language toolkits.
As for his complaint about nvim. Compiling this from scratch is trivial for vim and neovim. It is a 5 minute job. I would agree if it is large applications or suites, but for basic tooling it is a non-issue.
The biggest problems with Debina IMO is that they insist on having too much shit installed by default (e.g. games meta package get bunged into the default desktop) and they fuck around with packages toon much (e.g. what happened with KeepassXC last year).
He also rated FreeBSD highly and I've never had much luck with it.
I'm putting this in the Linux Thread because it's been discussed here the most:
The owner of paint.net (getpaint.net) actually got the domain paint.net after fighting domain squatters for 22 years. It's latest owner started hosting content about paint.net and deceive users and with the help of a lawyer (for trademark infringement/domain squatting), he got the domain:
(link / archive)
You can now visit paint.net and it serves as a redirect:
I'm putting this in the Linux Thread because it's been discussed here the most:
The owner of paint.net (getpaint.net) actually got the domain paint.net after fighting domain squatters for 22 years. It's latest owner started hosting content about paint.net and deceive users and with the help of a lawyer (for trademark infringement/domain squatting), he got the domain: View attachment 9072287 View attachment 9072291
(link / archive)
