A Future Without SystemD for Solus

Tux29

Hi everyone!

I want to open a discussion about implementing an alternative Init to SystemD for Solus OS. The most important thing of all is to give the end user a free choice from SystemD, always

brent

Tux29 want to open a discussion about implementing an alternative Init to SystemD for Solus OS. The most important thing of all is to give the end user a free choice from SystemD, always

I know some Distros have different inits you can choose from but I've never seen a distro offer a choice between systemd and init. I shall abstain from the vote out of precedent (or absence of).

ReillyBrogan

Tux29 There is literally zero chance that the Solus developers will switch away from systemd at this point. We are completely happy with systemd, why would we switch to something that is worse?

Original-Syn

brent MX and Gentoo come to mind.

Tux29
That said, that seems like a pretty big undertaking for the Solus team and would pull resources from maintaining and improving Solus as a whole. Would this be something that you are looking to implement and maintain? How would you see it being implemented to handle the established calls and functions that already exist in Solus?

I am not against alternate init systems, but I prefer systemd, even as an old Linux user. If it could be done in a way that would not burden the team or cause issues, by all means. If not, no chance.

Staudey

Tux29 The most important thing of all is to give the end user a free choice from SystemD, always

Why would that be the most important thing? It seems like a huge waste of time to support a second init system, more than anything.

(btw, it's not "Solus OS", just "Solus")

AlphaElwedritsch

Why do your ask that? What's the intention?
What's bad with the existing one?
I don't understand this thread without any background information...

Tux29

AlphaElwedritsch Because for years there has been this diatribe between the supporters of SystemD and those who would like something more inclined to the GNU philosophy. I'm more in the middle, also because I've never tried any Init other than SystemD.

Honestly, knowing that 99% of the most popular/most used Distos are all SystemD approved makes me turn up my nose. I think that in some of them, I'm thinking of Fedora or Garuda, or even Solus itself, there should be community Sipn that include Init that is NOT SystemD; and be able to try and use them daily.

brent

Original-Syn MX and Gentoo come to mind.

yay. I can vote now 😉

Tux29

Original-Syn with the era of "Universal Apps", "Virtialization" and "Containerization" increasingly present in recent years, I think independent Distros should increasingly focus on something that differentiates them from the rest of the crowd. At the very least, ensure that there are community Spin that do so.

AlphaElwedritsch

And so the vote leads to nothing and is merely hypothetical. But it can continue to fill the summer lull. As a pastime... 😉

alfisya

https://blog.tjll.net/the-systemd-revolution-has-been-a-success/

CorvusRuber

Tux29 I want to open a discussion about implementing an alternative Init to SystemD for Solus OS. The most important thing of all is to give the end user a free choice from SystemD, always

Usual question: why ?
More specifically: why, on a curated rolling release distro, "free choice" should be more important than quality ?
Secondary question: do you have an idea of how much work will this require ?
The Solus team is pretty small, there's simply not enough manpower to redevelop (and retest, and redeploy) the whole distro simply for the sake of decoupling from SystemD.
Last questions: are you an active contributor to Solus ? Do you financially support the distro in some way ? Because if the answer is "no" in both cases your "discussion" starts sounding a bit.....entitled.

Staudey

Tux29 I think independent Distros should increasingly focus on something that differentiates them from the rest of the crowd

I mean, sure, but it should be something that also makes them better, not just different for the sake of being different.

TraceyC

Tux29 I think independent Distros should

It's easy to tell other people what "should" happen when you're not doing all the hard work to make it happen.

If you're truly interested in what distros "should" do, donate your time and help make your ideas reality. Otherwise, what right do you have to tell other people what they "should" do with their time and energy? This does sound entitled, indeed.

While we're at it, I think someone should fund a luxury vacation for me at a beach resort.

alfisya

The staff are not interested in this idea and that is all.

Many of our tooling are in the open, everyone is welcome to do whatever with it. Fork it. Break it. Build something with it.

Citfta

I have been a Linux using for only about 5 years now. But I have tried many different distros. I have joined forums for most of them. I have heard all the arguments against systemd and some for systemd. When I discovered Solus around the first of this year I was very surprised at how fast it loaded and how well it worked with technical programs like Orca and FreeCad etc. Only later did I do enough digging into Solus to discover it uses systemd. I thought to myself then "Why all the negativity about systemd. Solus works great with systemd."

My opinion is that if someone has a dislike for systemd then go find a distro that doesn't use it. Change, just for the sake of change, is usually not good.

Original-Syn

@Tux29 I would like to mention that I do appreciate you thinking of ways that you believe could benefit Solus. I know we may not agree with the idea, but I think it should be encouraged to put out ideas that the community has.

From some of what you said about newer and/or smaller distros standing out, I think that is something that Solus actually does well. It is an efficient, curated distro that is rolling, but at a pace that makes it more stable. It is not just another retread of Debian/Ubuntu, Fedora, or Arch with just some fresh coat of paint and changes here and there. It is its own independent distro.

While I think it is always good for users to have choice, that does not always line up for every distro or piece of a distro. Something like the init system which is truly a core piece of the distro working along side the Kernel, that is not an easy task. You have to build the proper abstraction to decouple the current init system to do so and provide the interaction in a way to allow for it properly. When you are talking about systemd, you are going even further in that you have a lot of the current and future packages and even desktops (see Gnome) moving to systemd only support. That means you would have to ensure 1:1 compatibility. That is a lot of work for any team, let alone a smaller team on an independent distro. I have worked and contributed to Linux and FOSS for over 3 decades, it is not an easy task.

Finally, while there are pockets of users that are anti-systemd, many from my generation that started in the early 90s on Linux, it is getting smaller and smaller. Even if you look at a current post on Reddit about the article posted above, you see that it is very much one-sided in favor of systemd. I, personally, don't mind using non-systemd distros if I desire to use that distro. However, I do not see a strong enough case for the dev team to spend a ton of time to implement and then even more time split between supporting and maintaining multiple init systems.

So again, I do appreciate people wanting to bring ideas to the team here and just to be clear I am not on the team here, I am just a supporter of Solus. As mentioned if it is something that you are someone wants to look more into and implement, the beautiy of FOSS is that you can fork and re-imagine it to what you desire to see.