This. Not only does it save space, but you develop startup scripts using something people have tested the crap out of on a daily basis.

5 members found this post helpful.

You have to say his name 3 times.

1 members found this post helpful.

@glorsplitz, look away now!

No, these are messages early in the process that for some reason dmesg doesn't log and is long before syslog takes over. However, this experience is from some time ago and I don't remember the specifics. I just remember there was something that caught my eye during the early stages of booting the system, and when I went to go find it, I found it wasn't in dmesg (or any other logs).

No worries there. I don't let situations like this come between any of my friendships. You still provide valuable help in the forum and I greatly appreciate it and you. I feel the same about Darth, Drakeo, PROBLEMCHYLD, and a few others I get into regular disagreements with on the forum. I still respect them as people and even if I don't agree with everything they post, they do/did provide valuable help on the forum and I don't wish any of them to stop contributing. And I'm not so high and mighty that I know I am right in all situations. If I post something that is wrong, questionable, or weird, I hope people call me on it.

As for the little or no benefit in the Slackware subforum, we can surmise that this is not going to be entering Slackware anytime soon, so I don't understand why posting it here is appropriate vs another more generalized forum. That can get you the people with experience with it and would likely weed out many of the anti-systemd-in-Slackware people in this forum.

It's kind of the point that I'm REALLY glad systemd is not part of Slackware and not likely to anytime soon. That a BSD guy would see value in it made me wonder if I am just set in old ways, knee-jerk resistant to taking on new learning curves (on systems I've used I learned just enough to get by) , not to mention having arrived there as part of an extremely polarized and heated explosion of "which side are you on?" with much of that fueled by cult of personality, and as a result, missing something. This concern has recently been exacerbated by a huge amount of headaches and hard work from Pulseaudio that basically blindsided me in 14.2 and while I'm intensely grateful that 15.0 or whatever the next full release will be called, will come with a Pure Alsa option but with the "feature creep" of Pulse extending to apps far more fundamental than Bluetooth, who knows how that battle will end? If there's a possibility of getting blindsided again, even in 10 years, I want to be prepared.

I don't like it in myself when I arrive at conclusions during heated arguments. They are rarely reliable and often foolish. I don't want to get blindsided again but even more important is not in effect blindsiding myself by painting myself into a corner or cutting off my nose to spite my face.

For these reasons I posted first in the Slackware sub-forum hoping to make more solid my conviction that systemd is still a bad tradeoff while viewing it in a less heated light and at the same time at least explore the faint possibility that my assumptions during a heated flame war were not axioms and also that the maturity of 7-8 years experience in other distros may have improved systemd in a way that could shift the cost/benefit profile. While a scant few benefits have emerged of which I had been unaware, and it matters that it is in this sub-forum and not one that is locked into systemd, I don't yet see any appreciable shift and in fact more actually experienced problems have been mentioned than benefits. Nobody has said "I was forced to use systemd and at first I hated it but now I can't imagine working without it. It makes my workload easier". Maybe that will occur in more open sub-forums that include a lot more devotees. I doubt that but I would prefer to know.

I'm going to mark this thread "solved" since it is unlikely any earthshaking news will occur. The answer to the question posed in the thread title seems "The flames have died down substantially but the coals are still hot and you need to run pretty fast to get through"

Thanks everybody for your participation and input.

EDIT: BTW I was unaware there actually was a "System D" of French derivation (from "débrouille", débrouillardise or démerdewhich) that could have caused some confusion but I strongly suspect that may be how systemd got it's name in the first place.

2 members found this post helpful.

(from "débrouille", "débrouillardise" or "demerdenzizich" ):
The last word is not exactly french it's a mix of french and german : "demerdenzizich" and not "démerdewhich"

To be honest, the above link sums up the issue in my mind. The links provided therein appear to be fair.

YMMV.

2 members found this post helpful.

The s6 supervision suite A modern alternative to systemd
https://www.youtube.com/watch?v=gZqIEstv5lM

For those who want some quick introduction to init systems (the video duration is around 15 minutes and although the title is about s6 supervision suite, it also talked a lot about init systems in general along with some examples of available init systems out there).

I thought some people might be interested or find it useful.

4 members found this post helpful.

I'm quite sure you know that but I think saying that 'dmesg' logs something can confuse beginners. dmesg does not 'log' anything, it's not a daemon but a relatively small program that uses libc klogctl() that in turns call syslog(2) to dump kernel logs. On Slackware we also have klogd that in turn is a daemon and it really intercepts kernel logs as you can see in /proc:
and passes them to syslogd so that you can see them in /var/log/messages.

systemD is too windows-like for me, please no

1 members found this post helpful.

How much more time to do you need to make an "intelligent assessment"? It has been how many years, and been tested on how many machines, in how many situations? Any assessment that required intelligence was done long ago and now is just the waiting game to see how many more times original assertions are proven right. systemd will die sooner or later because it is flawed by design and so was sysV. Who would have thought it would have lived 20 years beyond it' s time? It did. It died. So will systemd.

Stuff I looked for in the thread but didn't find:

1. Was pushed through hastily during a stifled debate that leaned on straw man versions of every argument made against it.

2. Deeply arrogant condescending developers-- not just to a community that saw problems-- also to a point that it won the developers a Pwnie for their bad attitude towards security.

3. Openly aimed and described by lead developer as replacement for more things that people would rather keep, yet they're insulted for wanting to keep what actually works.

4. Could name several technical problems, why bother though? Trying to talk technical problems with systemd fans is like trying to argue with Jean-Luc Picard about "midichlorians" while drunk on the Enterprise holodeck. They have their own religion/cult based on technobabble borrowed from Microsoft, Apple and pointy-haired management types.

5. An init system hosted and developed on servers owned by Microsoft. Really, *that* Microsoft.

6. Years in, who honestly needs it? (This one was already covered but really rounds it out.)

7. Above all-- you would think it was Lennart's computer, not MINE, for crying out loud.

If I wanted any of this, I would have installed it. I didn't, I still can't figure out EXACTLY why the DD's did. Sure there's no shortage of theories and excuses-- but only that.

I'm used to Debian warning me if it wants to change TZ data. But when it quietly installed a new init system, it wasn't mentioned. Why isn't rc.local running? Oh, I'm the idiot now. KTHXBYE!

I left Windows behind to get away from every attitude, posture and practice behind this thing. I actually hate the attitude and the costs, more than Windows itself. But some people just don't care about that. Fine, I always thought Debian did-- and that's why I chose Debian.

I chose Debian for precisely the reasons that ceased to be relevant when that nonsense was introduced. But whatever, I'm a very silly person. Everyone that doesn't want this piece of garbage is just as silly and misinformed, or so I'm told year after year.

But reason #5 by itself would be enough for me.

4 members found this post helpful.

People have seemed to forget why Microsoft was so hated. The mass hypnosis of the masses seems to have erased their memory of Microsoft. 5 is a very good point. The purchase of Github was the ultimate coup Microsoft ever did. Embrace, Extend, Extinguish. Something Red Hat learnt very well. A side note: people worry IBM will swallow RH culture. I bet it will be the other way around. The whole campaign for systemd was a FUD campaign, and then RH pushed with EEE something that was straight out of Microsoft. There was obviously some backdoor dealing going on in Ubuntu/Debian to switch for no valid technical reason.

Now, every non-systemd distro I tried boots faster than systemd. Having a problem with gophernicus magically disappearing under systemd but still serving gopher. It is touted as secure, maybe one of the secure things to do is disappear and keep serving under systemd.

There is no actually no "intelligent assessment" of systemd the more I think of it. When people tried it turned into FUD flame wars against the heretics.

All complexity breaks down into smaller parts and the more complex it is the faster the entrophy sets in. Encourage systemd to expand, and extend, so it extinguishes itself (/tongue in cheek).

1 members found this post helpful.

I'll point out that the very nature of Git (a distributed version control system) means that the control of GitHub doesn't mean dick.

No, really.

My clone of a repository is as fully capable of being cloned by someone else as the version on GitHub happens to be. Christ, I've even repopulated an effed up branch on GitHub from a local clone.

MS bought GitHub to get the money paid by companies to (essentially) back up their private source code on GitHub. Anybody who works on open source can use GitHub or not.

Microsoft does not think on that level and is one of many reasons it is the dominant consumer desktop. You are not wrong but it is irrelevant.

systemd is open source, in many git repos outside of Github, and that did not stop it from dominance according to the three E methodology.

Microsoft has so much money the making of money is not a reason anymore. Money is a low level tool to them, that is all. Those without vast amounts of wealth think of tools to make money, those with vast amounts of wealth think of money as a tool to achieve their aims.

I believe this is a fruitless conversation but I'll simply point out that most consumers have absolutely no idea what GitHub is or what it does.

SystemD doesn't have a freaking thing to do with MicroSoft and has everything to do with RedHat. I'll point out that for quite a few years, if you asked someone what Linux version they were using, they would give you a RedHat version number.

Given that RedHat has corporate backing (not evil in and of itself), it has a lot of power to twist upstream projects into directions that it would prefer.

I work in an American company myself, and at least the engineers I've worked with spend approximately 0 hours a year scheming ways to bend the rest of the world to our view of how software should be written.

I suspect that the team members who wrote/support systemd do not believe that they are evil or doing anything intrinsically wrong. I suspect that they are engineers who believe there are trade-offs to anything that they do and that they have chosen trade-offs that maximize what they think is important.

OTOH, a non-empty subset of us have a set of trade-offs that is different from those the systemd engineers set.

BUT, it takes time and humans to write software. I know that I spend more time here than perhaps I should, but almost all of us work an ~8 hour day to pay our bills. The remainder of our day is spent with our families (as it should), sleeping, and doing other stuff.

A company that hires people to write software will almost certainly bury those who write software in their spare time.

2 members found this post helpful.