gotta say, i started this thread for those who have already run the gauntlet, found the information they need to make a decision, and had decided they dont want systemd... but since the thread's asking for info, and there's the wiki page now for those seeking how to avoid systemd, i might as well play...

utterly exhausted myself trying to answer that in person to too many people. thankfully, i was just reminded of this page: http://boycottsystemd.org/
...which along with the links in it, should help to paint the picture with some extra clarity on some details. (and likely does better than i can)

if that's still not exact enough, one can also go search for bug reports (across many distros and projects too), in which there are some highly technical minds ripping it to pieces, large and small.

indeed seems to be an argument without end at times indeed... tho i try to retain hope.

when you give in, and use systemd, you likely dont see a problem... from an individual home-user perspective at least.
when you try to live like a geek and learn your system, or try to exercise the freedom to choose [an alternative init], you start to see [at least some of] the extent of the problem.

like is subtextually hinted at in jpolard's post with the numbered list, it's all rather disempowering, in a similar way to being fed a fish, rather than being taught to fish.

in many ways, it's reducing the dennis ritchie-esq genius simplicity to a plastic-wrapped consumer product.

all that diversity born of the freedom, the GNU'ification of unix, is a wonder, and a marvel, and not something to be sacrificed for "the one linuxd". systemd in this regard seems to me like an innovation to hinder further diversity of innovation.

thanks for saying that. i too had hoped for a straight up practical help resource for the interested parties... i guess that was too much to hope for with the liquid tungsten hot topic that is systemd.

^_^ "boot slightly quicker". yep. that was the only real benefit i got out of it... and not really all that relevant considering my machines are "always on", and with classic inits they already booted a whole order of magnitude faster than windoze anyway. heh. too high a price for too little gain.

2 members found this post helpful.

I'm not bothered about shaving a few seconds off my boot time.
I also prefer individual tools to one of these.

a friend reminded me of this excptional article too. http://pappp.net/?p=969

... so we see, there have been cooler heads looking at this for a while too. indeed it can be difficult to find, or to even bother trying to look for though, given the sea of emotionally charged & strong worded exchanges and rants about it.

i really like the clarity in that article, depicting systemd as the polar opposite, in many regards, to unix philosophy.

1 members found this post helpful.

Off-topicness aside many of the comments made both for and against the usage of systemd as well as alternatives are all valid points as to the nature of the beast at hand... namely... the implementation itself thereof.

To be honest, we all knew as soon as udev was folded in and journald became mandatory the project was going to sidetrack out of just being an init replacement, and thusly it has.

I think systemd in it's original intention was sound as solely an init system, but when it started to bring in more and more subsystems and services, the project stopped being it's intention, and became something else it shouldn't have. Egotism undeniably drove this project to become what Christopher Barry, respectively, called it out as, a virtual kraken with many too tentacles.

I feel Christopher Barry is dead on the money with his post at LKML. Systemd does offer a lot, but the real question nobody has ever wanted to bravely answer is why is it reinventing wheels, some of which are nearly 13 years old, all in one dangerously seemingly convenient package? That question is constantly avoided by the pro-systemd crowd, and even some of the anti-systemd crowd.

We all knew GNU/Linux was designed around communal and collective efforts of many to promote the whole by it's individual parts. GNU/Linux promoted for the longest time, the UNIX philosophy. So why is there such as rush to abandon the ship when the ship isn't sinking, nor is it in danger of sinking, and only needs some paint, and a few new crew members doing tasks nobody was doing?

Was the original intention only to replace a functioning initisation system with an unneeded new one, or was transforming Linux into something else the original intention, with init only being the first step in the process? Considering the people behind the project, I highly doubt systemd began as a small idea and grew uncontrollably as the result of momentum. Which takes us back to the topic of the thread. How long can stop-gap measures keep systemd at bay? Will these methods continue to work as the tendrils grow deeper? In other words; are the methods offered of avoiding the beast long-term solutions or merely short-term work-arounds?

1 members found this post helpful.

I don't think the question has been avoided by anyone, I think people see it as yet another way to do things. How many init systems are there? Why do you think people need to ask this question now with systemd when, for example, it wasn't asked when Canonical created upstart? Why do we have so many DEs and WMs or Distros when in reality we only need 1? Why wasn't this same question asked when Gnome was started after KDE or other DEs started after both Gnome and KDE? The reason no one asked them then is because it gave people choice and believe it or not you still have a choice.

All this wailing and gnashing of teeth and comments of "questions are being avoided" takes away from the fact that you as an individual still have a choice. Exercise it and let others exercise it the way they see fit for them. Accusing people of "ignoring" important questions ignores the fact that you have a choice just as they do. Sure they may have chosen something you personally wouldn't have but that doesn't make them any less capable than you of making the correct choice for their personal circumstance. To me, so this is personal opinion, the entire premise you base this particular argument on actually belittles your argument against systemd simply because you appear to want to deny people the right to make a choice.

A big part of the problem is that the creators of systemd have done something very smart. They focused the debate on the init aspect. By diverting the debate away from the big picture, the result is most people seem to believe that systemd is simply another option for initialisation methods. Yes, there is choice. OSes are free to use whichever init systems they want. As long as the system remains modular. By integrating many processes, systemd removes choice, not increases it.

Oh wait. I forgot. I still have choice. If I do not want to use systemd, I am free to create my own OS: kernel and applications.

1 members found this post helpful.

You do have choice, there is Slackware, Slackware derivatives, the BSDs, and other Unixes. There is even Hurd. You are not being forced into accepting systemd. The majority of people doing the complaining are quite capable of exercising freedom of choice and are technically minded enough to adapt to other systems if their desire to stay away from systemd is strong enough.

Let's remember this thread was started to offer more choice, to offer help and encourage people to share what they have done to avoid systemd in their chosen distro (that may or may not already be using systemd). The discussion about the merits/drawbacks of systemd and the personalities of individuals involved in systemd is a side track that is in the myriad of other systemd threads.

indeed. something i highlighted in my original blog post about this. ultimately, the tyrant must be defeated.

i'm not sure you've been paying attention. that choice is under threat. ever increasingly hindered and coerced. the workload ever increasing for those who want to help retain/maintain/provide those choices.

it would be a non-issue, like you depict, if it were so simple as you say. it's not though. try living without systemd, using another init system, without giving up any of the tools you use, particularly on a distro that's decided to become systemd-centric, see how easy that is for you. ...and the direction its heading, seems set to increase the problem to great extremes.

dont let the tips at the start of this thread delude into thinking this is a trivial fix for the long term.

the situation is getting close to analogous to living under an authoritarian regeim, and someone saying "if you dont like it, move somewhere else". :/

Although I like reading about the methods people are coming up with to avoid systemd, out of curiosity, I have three concerns:
1) A work-around that by-passes systemd is not really avoiding it. The monster is still present. Some may, and probably will, disagree with me, but in my view it only adds another layer to an already overly complex system.
2) Such work-arounds are only temporary. As the scope of the systemd project grows to encompass more of the system, dancing around it will eventually become impossible.
3) At present three distributions are avoiding it, but for how long? Gnome is already moving in the direction of systemd-dependence. Soon, if not already, Gnome will only be able to run on systemd OSes. But when the rest of GNU's offerings follow suit? Not only applications like GIMP, but what about GNU utilities which are at the core of every Linux system? Can they be kept out of systemd's clutches? If, or probably when, KDE and Xfce follow Gnome's lead? It is easy to say no to one DE, but not to all of them. I would be happy with a system that only has WMs, but the developers of most systems would not be.

In short; I like the idea of avoiding systemd, but I doubt it is possible in the long-term without moving to BSD or Plan9, unless someone with Torval's ability makes an alternate Linux kernel.

Actually I have been paying attention and I maintain you still have a choice. Slackware, BSDs, etc.

This does not change the fact that there are still options. People left Debian for Slackware because of Debian's move to systemd.

This is Linux,nothing is trivial.

And with that comment I bow out of this thread because that comment was just over the top. I'm sorry if you, and others, believe this is such a drastic (as though it is a life and death) matter. I'm no fool, although many of you will think I am, just because I refuse to think systemd=lifetime enslavement to some evil overlord.

Unsubscribing.

One viable solution would be to get things like ConsoleKit, haldeamon, and other daemons and services that were forcibly deprecated, fork them into new projects, and keep them updated with modularity. This would mean eudev would have to fork eventually also to keep netlink intact.

The problem is, choice must be kept regardless of the situation, and when that choice is taken away we will not have distributions of GNU/Linux, we will have distributions of LinuxOS. Do we really want GNU/Linux to become another OS-X or Solaris clone? I doubt it, but that's the direction we're heading.

Regardless of the controversy surrounding this statement, "GNU/Linux has always been about choice." it is fact that choice gave lots of innovation to GNU/Linux, allowed many projects and clones to develop and thrive. If that choice is eliminated we will not have an environment where choice can thrive. The same development and innovation stagnation the systemd developers preached about with other UNICES is all they are going to promote on GNU/Linux.

If this ends with GNU/Linux and choice, then it's a win for everyone, but if this ends with LinuxOS, we will loose greatly, and honestly we'll have nobody to blame but ourselves.

i'm rather amused that "getting close to analogous" (and the context of the end of the sentence which carried the point) seemed to be overlooked, and "authoritarian regime" was bumped up to a more specific and extreme "lifetime enslavement by an evil overlord". ^_^

anyhoo, in other news...

the uselessd camp are doing their thing now (yay, a fork of systemd from before it got too over-reaching), and there was even a nice article by one of their number that made mention of a few of the forgotten inits. i'm sure i'm not the only one inspired to keep digging into the positive side of this, and letting it spur on further investigation and potential for innovation. despite whatever magpie-isms & capitulations are happening with systemd becoming a dependency of more n more of our software elsewhere, the geeks go marching on.

1 members found this post helpful.

What is wrong is the pact between Red Hat (systemd pusher, as well as udisk another piece of poor software) and Microsoft.
https://www.redhat.com/promo/svvp/
http://www.eweek.com/c/a/Linux-and-O...nection-Furor/
Microsoft will never give up on destroying Linux and having back door software patented by Microsoft is one way to go. they will pull the plug when it suits them.
Never trust Lucifer or any one in league.

1 members found this post helpful.