Interesting...maybe. My understanding is that LP was not told to develop systemd but exactly the opposite but he did it on his own time and dime. It has been realized a long time that LP is fascinated with Apple OSX and wanted to unify Linux along those lines which just happened to fit into Enterprise ideals. I had no idea that RH was a business failure and I'm not certain that is true. Can you cite a source?

Anyway my point certainly isn't that LP is some sort of genius but rather that it should be obvious he's no dummy either. Now that I'm all but fully retired his goals are not at all consistent with mine. I'd prefer Linux remain more a hobby OpSys, loose and funky but extremely creative as it has been. I rather hate the idea of it even possibly becoming locked down in any way. It remains to be seen if systemd will actually increase that likelihood or not.

Ivandi cease and desist immediately or I will cut you right out. You add absolutely nothing of value let alone on topic and seem to be trolling to create a flame war. Stop now please. There will not be another warning.

So I got curious and looked it up: While RedHat is not as profitable as they were during the tail end of the dot-bomb era (to be fair, neither is anyone else in the linux arena), since then their pre-tax profit margin has been fairly steady, hovering around a very respectable 15%, and they've been steadily reducing their long-term debt. (For those not aware, most larger businesses run on credit.)

For comparison, the profit margin of the average successful grocery store is about 1%, and IBM hovers around 11%.

RedHat 2006-present:
https://www.macrotrends.net/stocks/c...-profit-margin

2008 - employed by RH
https://en.wikipedia.org/wiki/Lennar...ife_and_career
2010 - as a RH employee - here boy, a bone to play with, we need some sort of systemd (not RH branded) to gain popularity & competitive advantage (see my post #436)
https://en.wikipedia.org/wiki/Systemd#History

Regarding RH, the "business". A business, as expressed in economic terms, is actually an endeavor/enterprise/undertaking/venture which is usually initiated based on a few pillars: vision, mission and values.
Without diverting this thread too much & wasting space, here a short summary that I found pretty good:
https://courses.lumenlearning.com/wm...on-and-values/
If you end up selling it, you actually failed in developing and sustaining (believing in) your pillars. It's that simple, no citations needed.

So, Pat packages software that is given away for free into a distribution (which we all are entitled to do) and instead of just keeping it for himself he gives it away free of charge, and also gives people the option of donating money. How this is twisted into a negative is beyond me.

He contributes Slackware which has served me and others quite well over the years. But if you're talking about money how do you know if Pat donates to software projects or not?


Prove it.

You do realize that you're one of the most grumpy people on here, right? ;)

To be fair, the very existence of this topic is trollbait. Your intentions may have been innocent in starting it but systemd is, and will be for the next generation or so, a divisive topic. If anything, I would request a thread lock to put it out of its misery: it's already marked as solved.

Trolls don't require bait. They pop up everywhere they think they can get away with it. IMHO we can't let that prevent us from discussing controversial issues or we get stuck with just discussing the weather and what we ate for lunch. I know it's marked "Solved" because I marked it precisely to point up the conclusion that many cannot rise above their prejudices and echo chambers to discuss anything without ranting the same old crap we've all heard/seen a million times. It's not flames now. It's just boring and ignorant (not you... the wannabe flamers). I'm not going to request a lock. I intend to report anyone who goes off the rails to the degree that ivandi is currently flirting with and judging from the responses here I doubt anyone who crosses that line will be happy with the results.

I consider it a valuable privilege to be a part of a rare, open, adult forum such as LinuxQuestions. Anyone who doesn't support such responsibility and prefers to act like a petulant child needs to leave it for those of us who revere such earned respect.

In my testing of systemd distros on bare metal (Debian, Arch) I found that systemd had some benefits, that is, for the most part both distros shut down very quickly when the shutdown command was issued. Both distros were reasonably stable (Debian under KDE-plasma had some lock-ups with thunderbird). I found the use of systemd commands to be at first confusing and annoying, then acceptable. In my opinion the notion that systemd is a cluster#@*% init system is a bit overstated. I think systemd does what it is designed to do reasonably well. If upstream development forces our maintainer to consider systemd that won't be the end of the world, Slackware will run just fine with systemd. However, at this time systemd is not something that is necessary for Slackware(we don't need it). That init system isn't something that we'll need to fret about for several years. I'm still ambivalent about the notion of Slackware having systemd, that is, our maintainer will make the call. It's all good. Praise Bob.

But your handful of desktop PCs is not a big enough sample. ttk and his/her colleagues manage or develop on 2,000 servers, and they are constantly struggling with systemd issues. The verdict of people who manage servers at this scale was a warning for me. Another warning came when Red Hat introduced systemd into their so-called enterprise operating system (which is increasingly just an agglomeration of acquisitions) before it was out of beta. I think work started on systemd some time in the late 00s and by the early 10s RH had slipped it into their enterprise OS. Who does this with a technology so critical to an OS as an init system? Most critical GNU/Linux software has been battle-tested for years, some of it for many decades. systemd went in straight away, and was adopted by 90% of the Linux lemmings soon afterwards. That suggested to me somebody has a vested interest in seeing this obfuscated code foisted onto the software the entire human race will be using in the not-too-distant future.

Yes. That's certainly true; I was speaking from a limited use case (my desktop). Our friend ttk is better able to speak to systemd issues in production. I am very grateful for Slackware. :)

It is a good summary of what I hear when talking with (professional) sysadmins:

- young admins with little previous experience find systemd rather easy (and find the maze of system V init scripts rather klunky and/or intimidating). The fact that whatever the question they have about the systemd way, they find online answers and tutorials everywhere, certainly helps.

- older admins find it "at first confusing and annoying, then acceptable.", as you put it.They did not _choose_ systemd, but were forced into it, because RedHat/RHEL (which is synonym of Linux in their IT dept., most of the time).

They may grumble at first (remember old means grumpy, if I follow some of the forum members), then they adapt and usually find it does the job. Actually they grumble mostly just because it is different, and they have years of "system V" muscle memory in their fingers. They actually don't grumble too hard because (1) like the younger admins, these days, they quickly find answers online when needed, and (2) because many of them already went through several init systems (eg. Solaris SMF) and various corporate tools to manage system V init scripts -- after years of proprietary UNIX, good sysadmins adapt quickly!

I didn't hear much "constant struggle" stories as reported by Gerard Lally. Most of the time, issues looked like trying to do things the old way, or just lack of familiarity with the new tools. Some early bugs have certainly be disruptive but (1) RedHat does actually pay people to fix these bugs, and (2) any new subsystem does introduce new early bugs: Moving from linux 2.4 to 2.6, XFree86 to Xorg, kernel video modsetting, etc... All may have been disruptive for some, but we wouldn't go back today.

I believe the biggest issue is not technical: It is that today, by large, young guys come to Linux and learn about the systemd way (or they won't ever care about their init system...). So when it comes to selecting a distro, what will they choose?

I haven't seen many grumpy people in this forum at all. Sneers, yes - we had one just yesterday. But grumpy? I don't think so. Contrarian perhaps, but holding contrary views about the init system is what motivated Poettering to start on systemd in the first place. If it's acceptable for him to hold contrary views then it's just as acceptable for his critics.

@philanc

Well, from my PoV, the biggest issues are technical. It's an invasive, inflexible, to some extent badly designed complex system that has no meaning other than to monopolize (RH & other "enterprise" distros) and ease the path for inexperienced - as Gerard Lally called (and I confirmed & elaborated) - bean counters. But you did mention these facts already. What strikes me is that you got yourself into the subjective critique, considering some members as grumpy. Remember, this is Slackware:
https://en.wikipedia.org/wiki/Slackw...ign_philosophy
Interesting is also your view on the "young admins" educational requirements, considering the ultra-simple, solid, flexible and transparent system V init scripts complicated. You inadvertently define these younglings incapable/cheap bean counters :) ... which is pretty much true, it's not fresh meat required anymore, but cheap meat.
I'm saying this because when I started using Linux (FreeBSD too) I had to learn (on my own) not only the init system but also the tools a *NIX system provides, in order to be able to admin it competently. Later in my professional career I also got a formal Linux Admin training (didn't ask for, was just compulsory) and again I had to learn (refresh) all the *NIX tools, including bash scripting, in order to get the certification.
Last point:
Yes, I mentioned that in some older thread/post. They blindly follow some documented procedure, without investing any effort in understanding what they're actually doing, calling the support when they get into trouble. And then everybody can write a cheap Linux Admin Guide / Training based on systemd stuff, it's totally democratized, idiots have rights too ;)
A rather sad development I might add.

Sorry for my poor wording. I don't mean people here are grumpy. It was just intended to be an (awkward) reference to Ivandi's post #464 above ("Pat is paid less than a seasonal mexican worker to compile the least functional version of LFS for a bunch of grumpy old men."). I am actually old, and many people would say I am grumpy when they have to listen to my rants about technology. Maybe I resent that :-)

Back to systemd and abga post:

I really appreciate minimalism, so I strongly dislike systemd. I agree with "invasive".
I don't really agree with "badly designed". I think Lennart Poettering is a smart guy (whether we like him or not, his track record is impressive, and I think systemd's design is rather clean _for what it does_. Of course the minimalist in me cannot like it.)

My point was that the overall systemd issue is not technical, ie. whether it is good or not. The issue is that it is here, and almost all over the place (redhat, suse, debian/ubuntu, arch, ...). So, like it or not, it is here to stay. Which means young people coming to linux will expect systemd exactly as we expect binutils, bash and gnu make today.

Look for example at bash: Years ago it might have been a question. Between the Bourne shell and the C shell, people could have argued that technically, the best shell was csh. The technical merits didn't matter much. The Bourne shell then bash won, and no distro maintainer would attempt today to keep all their scripts written in csh. Most (if not all upstream have adopted the Bourne shell (bash, or the painful dash). And how many young linuxers are today fluent in csh?

Maybe Slackware becoming The-distro-without-systemd could be a valid niche strategy?