It works great for me. I've been using Linux since late 1996, and one constant has been that the kernel and its surrounding software is always evolving.
People with legitimate use cases against systemd shouldn't use it. People who don't want to use it should use distributions that don't include it. But systemd doesn't need to "die", open source is about freedom of choice. People who zealously hate systemd need to chill out.
The biggest problem people have with systemd is that it's constantly growing, taking on more functions and becoming a dependency of more software. People joke that some day you won't be using Linux anymore, but GNU/systemd, (or as they've taken to calling it, GNU plus systemd) because it's ever-growing from a simple init daemon into a significant percentage of an entire operating system.
People worry that some day, you won't be able to run a Linux system that's compatible with much of the software developed for Linux without using systemd. Whether that's a realistic worry or not I don't know, and I don't really have a horse in the systemd VS not-systemd race, but I can appreciate being worried that systemd might end up becoming a hard requirement for a Linux system in a way that nothing else really is - you can substitute GNOME for KDE, X11 for Wayland (or Mir, I guess), PulseAudio for PipeWire and most stuff will still work, so the idea that systemd could become as non-negotiable an element of a Linux system as the Linux kernel itself rubs people the wrong way, as it functionally makes Linux with systemd a different target platform entirely to Linux with another init daemon.
Honestly, systemd isn't bad, because a one concrete program will always be more reliable than bunch of bash scripts tied with rubber bands and bubble gum, but Poettering is a twat by making it (purposefully) non portable.
It's opensource, you can port it.
The point is that he explicitly makes this hard. That's why he is a twat. The issue is that some applications (especially graphical) do get heavily integrated with it which makes it also hard to port them.
Wasn't there also a program that had exactly this goal? Making systemd more modular and portable and so on?
That said, having a common ground to build on is not bad and when you need tight integration. Sometimes you don't really want to do it twice. That's part of why Linux is so successful. It's a monolithic kernel that experienced widespread adoption and driver development. You really, really don't want to redo it all, if you don't absolutely need to. You really don't want to add complexity by making it all super modular either.
Pottering seems to be not exactly the nicest person, but he has a point when he says that one good solution is all that's needed and sometimes even better than two good alternatives.
Couldn’t leave you hanging…..
Not exactly an advanced Linux user but what's the hate with systems?
IIRC, it goes against the OG Gnu/Linux philosophy of having multiple small tools that interconnect well together. Systemd is seen as monolithic and cumbersome by some/most(?).
If someone is more connected to this debate feel free to correct me!
That's pretty much it. Systemd goes against the philosophy of "do one thing, and do it well" by doing a whole lot of things and being integrated to an extent that makes it pretty much impossible to use only an arbitrary subset of its components while replacing the rest with alternatives. I understand where the critics are coming from, but I honestly don't care either way.
SystemD is an init system (it initializes the userspace) and takes care of managing services. Its quite an important piece of software. There are many reasons linux elitists hate it . I honestly don't care about it as long as it successfully boots my system which it does. In fact I like the syntax for systemctl...
Obligatory I use Arch btw!
Honestly I actually like systemd, it seems to make more sense than those old initscripts to me.
I love systemd. Anyone who has tried to write initscripts or manage dependencies will agree with me.
Bzzz
You help where you can
