Linux Voice Wants To Know Why Discussions of SystemD Become So Heated
Linux - GeneralThis Linux forum is for general Linux questions and discussion.
If it is Linux Related and doesn't seem to fit in any other forum then this is the place.
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
Distribution: Red Hat Enterprise Linux, Mac OS X, Ubuntu, Fedora, FreeBSD
Posts: 89
Rep:
Quote:
Originally Posted by speck
Maybe it would still be possible for distributions with different philosophies to exist even with a wide adoption of systemd, at least hopefully that would be the case.
All unices can trace their lineage back to AT&T unix.[1] If you don't like the direction we're going you can fork, as has been done for the last 45 years. We have android, that's a fork, there is no reason why we can't have additional forks. However, you're in the minority, so the burden is on you if you want to fork. You have to weigh the pros and cons of forking, and in the end I don't think it's worth it. Red Hat Enterprise Linux has systemd, and since my career revolves around that product, I have to get on board the systemd train regardless of how I feel about it; in the end it's just an init system and is not worth bucking the trend to me.
If you don't like the direction we're going you can fork, as has been done for the last 45 years. We have android, that's a fork, there is no reason why we can't have additional forks. However, you're in the minority, so the burden is on you if you want to fork. You have to weigh the pros and cons of forking, and in the end I don't think it's worth it. Red Hat Enterprise Linux has systemd, and since my career revolves around that product, I have to get on board the systemd train regardless of how I feel about it; in the end it's just an init system and is not worth bucking the trend to me.
Unfortunately I think what you say is correct but I don't know how easy it would be to keep a non-systemd Linux alive (if other important core components and applications rely on systemd, they would need to be forked/rewritten as well). If it were only an init system then it probably wouldn't be too difficult, but we'll need to see how far systemd eventually extends into the Linux ecosystem.
Additionally, making things less configureable will reduce complexity for system administrators, unit file are simple to maintain, we get to offload the work of service management onto the developers.[/URL]
Having administered a number of client networks for a few years, i never found configurability a drawback, the above sounds like the thinking of the average software developer who thinks that his/her final version comes out pristine, perfect and able to take on all circumstances.
Thats just not true. The less configurable it is the less useful it is. out in the wide world there are always things the software developers didnt and couldnt take into account and the creators of unix knew this and created it the way it is, as small components that work together, for exactly that reason.
It seems to me that if the kind of thinking that came up with systemd had any practical validity, unix and linux wouldnt be the way the way they are today. I dont think systemd will kill linux or be unusable, but i think its highly likely that it will create the kind of admin headaches that its PR says its going to solve, due to lack of practical flexibility.
Quote:
Men are born soft and supple;
dead, they are stiff and hard.
Plants are born tender and pliant;
dead, they are brittle and dry.
Thus whoever is stiff and inflexible
is a disciple of death.
Whoever is soft and yielding
is a disciple of life.
The hard and stiff will be broken.
The soft and supple will prevail.
Distribution: Red Hat Enterprise Linux, Mac OS X, Ubuntu, Fedora, FreeBSD
Posts: 89
Rep:
Quote:
Originally Posted by fogpipe
Having administered a number of client networks for a few years, i never found configurability a drawback, the above sounds like the thinking of the average software developer who thinks that his/her final version comes out pristine, perfect and able to take on all circumstances.
Thats just not true. The less configurable it is the less useful it is. out in the wide world there are always things the software developers didnt and couldnt take into account and the creators of unix knew this and created it the way it is, as small components that work together, for exactly that reason.
It seems to me that if the kind of thinking that came up with systemd had any practical validity, unix and linux wouldnt be the way the way they are today. I dont think systemd will kill linux or be unusable, but i think its highly likely that it will create the kind of admin headaches that its PR says its going to solve, due to lack of practical flexibility.
It sounds like you've never had to work with junior administrators. I worked on a team that was responsible for over 5000 unix servers, and we simply didn't have the time, or manpower, to deal with edge cases like the ones you're describing. Everything we did was canned vender solutions. This is how enterprise IT works. Solving everyones needs is a long tail that follows the power law. The pareto principle and the principle of least effort also applies here. We can't solve everyones problems with one solution.
It sounds like you've never had to work with junior administrators. I worked on a team that was responsible for over 5000 unix servers, and we simply didn't have the time, or manpower, to deal with edge cases like the ones you're describing. Everything we did was canned vender solutions. This is how enterprise IT works. Solving everyones needs is a long tail that follows the power law. The pareto principle and the principle of least effort also applies here. We can't solve everyones problems with one solution.
Yeah i know, the systemd way is better because its monolithic, vendor approved, limits choices, will slow the development pace so that large vendors and their marketing droids can keep up, and bottom line, you guys really do know best for us i guess.
What was i thinking. Of course you guys are right.
Quote:
We can't solve everyones problems with one solution.
Ok now that part confused me. Thats exactly what systemd is trying to do and further more trying to solve a problem that doesnt exist.
The entire internet is running on linux and doing fine. Who needs systemd? Nobody. Except as you outlined above, large vendors out to cut costs. So i guess we have to put up with the drawbacks because redhat needs the money? I realize this had to happen though. Business interests often sacrifice choice and usability for standardization. I guess its time for something else. Linux was fun while it lasted, but systemd and the mentality that it represents will, i think, be the end of it.
Distribution: Red Hat Enterprise Linux, Mac OS X, Ubuntu, Fedora, FreeBSD
Posts: 89
Rep:
Quote:
Originally Posted by fogpipe
Yeah i know, the systemd way is better because its monolithic, vendor approved, limits choices, will slow the development pace so that large vendors and their marketing droids can keep up, and bottom line, you guys really do know best for us i guess.
What was i thinking. Of course you guys are right.
Have you ever heard of the Paradox of Choice? Linux is the fastest moving software development project on earth, honestly it does need to slow down for venders and end users to keep up. As a system administrator, I love the RHEL approach, I can do auto updates without worrying that my systems will break.
Quote:
Ok now that part confused me. Thats exactly what systemd is trying to do and further more trying to solve a problem that doesnt exist.
The entire internet is running on linux and doing fine. Who needs systemd? Nobody.
to enable multi-tenant systems by leveraging cgroups and containers.
for process, service, and socket management.
for transactional fault tolerant sockets.
for secure logging.
for faster reboots to minimize service outages.
to offload the maintenance of init scripts onto the developers.
You apparently know little about what systemd brings to the table, so I urge you to watch this video. There are also desktops, notebooks, tablets, phones, and embedded devices out there that can benefit from it, and these systems eclipse servers by a wide margin. Furthermore, the number of users exceeds administrators by a large factor. You have to look at systemd as a pareto improvement.
Quote:
So i guess we have to put up with the drawbacks because redhat needs the money?
What drawbacks? You haven't outlining anything that I would consider a drawback.
Quote:
I guess its time for something else. Linux was fun while it lasted, but systemd and the mentality that it represents will, i think, be the end of it.
Goodbye, if you don't like it you have the ability to fork it.
Pretty much bull pucky to this free living linux using biker.
Choice makes you unhappy? It won't sell around our campfire.
We'd pass him a beer though.
Quote:
I need systemd:
Alrighty then. I don't.
I need all cars off the road. They clog the streets. Motorcycles and bicycles are all that should be allowed.
But I guess that's different. Though the trees and vegetation might benefit from my line of thinking.
I posted my thoughts in the discussion link in the opening post of this thread.
In my uneducated way.
Systemd does not play well with others. Nuff said.
C'mon, fellas ... ... this is not the Faith and Religion Mega-Thread!
"nbritton" speaks from pragmatic experience, and no one can, really, "tell him 'no.'" If you have thousands of computers to deal with, as a great many of us do, then a lot of things have to be different from "what worked for a PDP-7 in a lab long ago." The great thing about Linux is that it is capable of embracing that sort of thing. You don't have to put up with any "one" solution to anything at all.
I just "shook my head in disbelief" when the "discussions" aforementioned very-obviously did "become heated." Never did figure out why that particular teapot had such a tempest in it. Never did try to find out. (And, do not want to hear about it now.) The incredible beauty of Linux is that, if you don't want to use this-or-that, you actually never have to.
I think its related to jobs, since systemd does so much more things at once, "all in 1 program instead of a lot of programs", and most of the changes are done by developers instead of administrators, why do companies still need so many administrators? They will most probably sack them and save money. If they need anything, they can just ask the developers, its not like you have to pay systemd developers if you want to talk to them.
The incredible beauty of Linux is that, if you don't want to use this-or-that, you actually never have to.
Sorry to drag you into this, but I think that's the crux of the issue. If it were only an init system, then I don't think you'd see anyone getting too worked up (similar to Ubuntu switching to upstart, no one outside Ubuntu really cared). systemd is already larger than just an init system and we're not sure how far it will eventually extend to (and potentially engulf/replace) other parts of what we currently call Linux.
Distribution: Debian Wheezy, Jessie, Sid/Experimental, playing with LFS.
Posts: 2,900
Rep:
I think its interesting that the people who are crying out (loudly) about standardising Linux have said they will jump ship, if they haven't already done so, to BSD because they want choice.
The 2nd sentence on OpenBSD's web page says "Our efforts emphasize portability, standardization, correctness, proactive security and integrated cryptography."
Considering this I wonder what is wrong with Linux standardising various aspects?
BSD was built on a singular system under a monolithic design because it was built to be a UNIX. GNU/Linux was built around modular projects coming together for a common design specification, but with flexibility in mind to allow multiple designs without an ironclad lock-in to anything specific.
A good example is systems that are embedded using busybox versus a standard GNU configuration.
"nbritton" speaks from pragmatic experience, and no one can, really, "tell him 'no.'"
NO! Please take the corporate "needs" elsewhere. If there is one thing that Red Hat has learned over the years, it is howto fork a community. Perhaps you missed the first round back in the nineties, the astroturfing is the same yet different because the psychology has seen an upgrade. ("Old code" serves no purpose, no giant shoulders exist). RHT goes on their way to the next quarterly report, and soon the world gets tons more cloudy crap and non-free OEM devices. Go on and cash in with APPL, GOOG, MSFT, AMZ et al as George Orwell spins in his grave, Some people must profit, no matter the cost.
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.