Systemd Hate Thread

...

Other urls found in this thread:

github.com/systemd/systemd/issues/5755
github.com/systemd/systemd/issues/5644
github.com/systemd/systemd/issues/6237
forum.voidlinux.eu/t/looking-for-a-systemd-free-os/2311/4
thenewstack.io/systemd-vs-linux-kernel/
github.com/systemd/systemd/issues/6237#issuecomment-312481714
gnu.org/software/coreutils/manual/coreutils.html#Disambiguating-names-and-IDs
twitter.com/SFWRedditGifs

The
fuck
Pottering

>Systemd developers have rejected mounting the EFI variables as read-only, since there are valid use-cases for writing to them.
Then stick it in a damned grub flag. FFS

One day we will have sensible init. But that day is not today.

what's the alternative to systemd?

There are a bunch. OpenRC is one.

...

> There is a use cae for that
Pottering will soon introduce DRM and webkit into SystemD.

OpenRC is sensible. Come to the Gentoo side.

>Systemd is replacing many os tools and components
>Supermarket and hypermarket is replacing local store and services
>tfw real life shitty things even happen in linucks

Link to the issue?

but OpenRC is just SysV

>Supermarket and hypermarket is replacing local store and services
Suitable analogy.

If you have an average or complex use-case, the availability of the tools you need increases and the cost of using them decreased with systemd.

For the people who only need a super simple init, try runit or something.

That reminds me, how functionally could you use the thing they had before that, with the /etc/init.d daemons
just wondering
for me systemd is just there most of the time, and then randomly starts raping my cpu and timing out because i somehow broke it by mounting things on a tmpfs mount i guess

OpenRC doesn't force you to run a specific init.

init starts OpenRC
OpenRC handles process supervision
init does nothing else but reap orphans

fucking Resolved is pain in the fucking ass

Everything becomes clear when you realize that guy is a government creation.

>yfw
>gnu/systemd/linux

That analogy is retarded, sysvinit will never be replaced with systemd like how supermarkets are replacing local businesses. For as long as forking software is allowed, systemd can't win.

Some context maybe?

Reminder that pulseaudio is also a work of poettering
Reminder that crap was also forced into users broken and buggy as it was
Reminder to unistall poettering crapware
The botnet called Avahi included

because everyone here reads the same github threads. fuckin retard

relevant: github.com/systemd/systemd/issues/5755

yep, that's the bother - makes load balancing and fail over a pain

did they ever put in the supposed backwards compatibility pottering promised?

>Sep 13
>"I have now locked this issue. I will unlock this again in a few days when things have become quieter. Thank you for understanding."
>still locked over a month later

tfw not-a-bug: github.com/systemd/systemd/issues/5755

Theoretically? Lots.

Realistically? None.

found it: github.com/systemd/systemd/issues/5644

void (runit)
alpine, gentoo (openrc)

>let's replace this working software with something entirely different that will takes years to get to the level we already had

Computing in a nutshell for about 10 years now.

how to start every service as root without being root: github.com/systemd/systemd/issues/6237

But if you pull down any one of the increasing numbers of programs that have systemd as a requirement, and systemd overrides everything else, then you've now got systemd.

That's why I said "Realistically? None" to highlight that in normal use-cases you'll eventually end up with systemd whether by design of the distro or just a personal mishap.

...

use flags

My feeling is that alternatives will emerge. Although most users will let it go right over their heads I think that the disgruntled ones represent a part of the community that just won't drop this or let it slide by. Something is going to come out of this.

Some systemd delete command follows ".." with ".*" and deleted some guy's entire fs.
Lennart in his usual ignorance defends that with that's how rm -r behaves too.
But it doesn't.

...

void sounds interesting

forum.voidlinux.eu/t/looking-for-a-systemd-free-os/2311/4

anyone know whether it's fairly active?

I use Debian with legacy sysvinit, and so should you.

Never missed systemd ever again.

"the issue is dead"

"the opposition to systemd has largely subsided"

"systemd threatened to succeed Mono cross platform .Net framework as the conspiracy theorists’ main target"

>itt, the tin hat crowd

thenewstack.io/systemd-vs-linux-kernel/

Thoughts on runit?

Very active. I got some help in the forums recently.

I have few questions about Systemd.
Does it compromises your privacy and security?
How good is devuan?
There is any alternative to Raspbian without Systemd?
How good is tiny core linux for the Raspberry pi?

>thenewstack.io/systemd-vs-linux-kernel/
>The Linux Foundation and Red Hat are sponsors of The New Stack.

I don't even have to wonder who could be behind this

fake news

Whyd they even switch over anyways?
like are there any advantages to systemd over sysvinit or what?

Because without it it would become too hard to package Gnome.
That was literally the only prevailing reason in Debian's debate.

Devuan is trash, do like me and use Debian+sysvinit, it's much better.

CIA and NSA paid them.

I̧͉̱̠̪̞̠͐̆̑͒ͪ̀ͅͅn̵̡̧̢̗̝̤̼͈͍̦̖̦̮̬͕̆ͮ̐͂̏̈͒̽̓͋̆̋͗͆̎̕s̷̜̭̖̙̹̱͇̜̠̦̦̬̳̘̭̬̤ͨ̍̈́̾̔͘͡t̶̢̫͖̩͓̖͙͕̾ͦ̆̂͊̂̚͜͞͠ả̞͚͈̜͎̳̳̘̦͕̪̠͕̩̻͚̥̘͈̋̄̏ͫͨ̓́͑̏ͯ̽ͤ̓ͨͣ́̚͢͟l̶͕̦̭̪̟̙̣̝̯̱͖̥͕̼̹͉ͯͤ̈́͋̂̂̓́l̡͚̘̪̟̗̺̰͍̹̈́ͧ̾̊ͧͤ̓͐͂ͦ́ͪ̓͂̌̉͆͢ ̡̨̳̙̺̘̦̠̞̂̔͌̄ͬ̄ͧ̀̔͒ͩͮĢ̧͓̬̜͍̙̭͔̞̥̒̉̍͐ͮ̑̆̽͆͑ͤ̔̂̔̽̿e̖̯̠̰͖̜̥͍͈̣̱̬̘̋̉̐̆̔̀͟n̸̯̪̘̤͑͋ͯͬͧ́́ͅt̶̸͖͓̟̳̬̞̼̪͇͙̜͚͛ͩ̽̔̐̈̕͢o̪͕̱̘̬̟͍̗͖͇̘͉͎̹̊ͧ̒ͯ͆ͬͨ̾͜o̤̟͖̹̬̯̬̖͍ͣ̊̄ͩ̓̿͋ͪͪ̆́͊ͯ͋ͯ̚̚.̵̢̝̤̫͚̜̰̳͈̃̂̿͂̎̍ͯ̍ͭ͘͞

This is right up there as an alternative. Debian has an enviable, if somewhat Teutonic, strain of conservatism. An obstacle if you're wanting the bleeding edge, but otherwise yea. And it's getting back feedback from all the distros it's spawned. I don't keep track of kernel development etc, so what do I know; I was surprised though when I found out that they had gone along with systemd.

yfw it was in the cards all along

I use it with Debian Testing (Buster), which is pretty bleeding edge being a semi-rolling release. (Now on Linux 4.13/GCC 7.2.0)

The "Debian is outdated" meme is starting to get tired, if people stopped using Debian Stable on desktops this meme would be over.

Because so many programs have systemd as a requirement that sifting through them to find those that don't or changing them so they don't, would be a huge waste of time.

Openrc and runit are both very sensible and perfectly good inits. Don't listen to the systemd shills.

Ok. Perhaps it's no longer true. Or maybe it was my insistence on using a stable branch. I can remember, reluctantly installing ubuntu in order to keep abreast of certain softwares I had to work with.

That's good.

I know this is probably typical of large software projects but this graph made me laugh. From systemd Github page

It's typical of hugely complex systems no one knows how to work with besides those on the inside.

But think of it this way: think of all the extra guaranteed wages that these programmers have given themselves.

It's not just Lennart and his cohorts.

Red Hat makes it money through support services.
Pushing complex critical software of which they have the most expertise in by far is a great business strategy.

Doesn't seem like a good long-term business strategy.

But with the decrease in cost also comes a decrease in quality.

Sure it is. Force systemd on everyone, make it very complicated and buggy, and now everyone will be calling you for support.

>needs superuser rights to start system services
>complains that someone with superuser rights can get superuser rights

>services owned by non-root accounts should start as root
hey poettering

>only those who work on the software know how to work on the software

STOP THE PRESSES.
Everyone in this thread is beyond retardation. The inane drivel posted in her is so ridiuclous, it's a wonder that someone can be so stupid not to realize they're THAT retarded when they post it.
To prove how predictable you are, I expect 10 posts using the words "shill" or other memes.

systemd shill

Prove how.

>non-root accounts
So every user then? There is only 1 root user, you fucking retard. Do you even know how the basics of the things you're talking about work?

Are you telling me that systemd is unable to execute a service as a regular user? How sad. Openrc can do this easily.


start-stop-daemon --start --quiet --exec /usr/bin/ -u USER

It can, you retard. Stop changing topics. Prove how a regular user without root privileges can start a service to run as the root user.

You mean
> X plus GNU plus SystemD plus Linux
>or XGSL

Or they'll switch back to Windows.

So you support vendor lock-in?

>vendor lock-in
It's free software, you fucking imbecile.

why we should avoid systemd?
Convince me to avoid it and tell me how to do it.

But the support you have to pay for to fix the intentionally broken software that is critical to your business isn't, is it?

So you stay locked in and pay the cuck tax, or you move to another system.

You are beyond retardation with your retarded internet buzzwords (cuck) and your conspiracy theories. You know nothing about software. You just like bikeshedding and talking about inane things to feel smart, but when someone asks you to show a modicum of technical expertise (like here ) you shut up.

You are another type of retard. The fact that you are still alive is astounding and you should be studied.

to be fair he's not wrong.

rm is just smarter than you and won't remove . or ..

blame bash and sh shells for expanding . .. when doing .*

It's not even the shell.
The wildcard expansion is done by systemd,
it's a delete command inside a systemd conf file.

>It can, you retard.
Not according to that bug report. Read it again moron.

>Or they'll switch back to Windows.
Windows server is 20x worse.

sensible maybe, functional no.

What function does openrc lack that an init system should do?

ok and?

same shit. the glob behavior is mimicking garbage shells. just so happens this delete commandlet wasn't as 'smart' as coreutil unlink()

HOOKS=(systemd autodetect modconf block filesystems keyboard fsck)


:^)

maybe have dependencies that isn't an incomprehensible mess of symlinks and stupid?

even LSB init headers are less shit than openrc.

That's not what the bug report is about, you fucking retard.

>it can't
systemd-run --user kill_yourself

>the brittle caps flinging autist

>tell systemd to start service as a normal user
>actually starts as root
Great software poettering.

It does what you tell it to.

Except it literally didn't. He told it to run as a normal user not root.

I have no idea what you're talking about. The only dependencies of openrc on my system are pam and psmisc.

>pam dependency
That's stupid, now it won't run on Slackware

are you retarded or play one on tv?

services usually have depends on one another.

It's the way it's packaged on my system. It's not required to compile it with pam support.

I still don't know what you're talking about. If one service needs another, you just tell it to start after it in the script.

The very concept of systemd is anathema to the original philosophy that has been so core to linux all along. The same philosophy that has been touted and retouted so many times over. Now that gets left by the wayside? The same philosophy that provided such a brilliant contrast to windows, back in the day, when people were getting sick of opaque monolithic mountains of garbage. Anyone who doesn't have huge qualms about systemd is either a kid who doesn't remember where the goodness came from or a retard. Mostly I suspect it's youngsters. Of all ages. I know why I began using linux, and this a quantum leap in the wrong direction.

It doesn't. Read this
github.com/systemd/systemd/issues/6237#issuecomment-312481714

If you still insist to know better than the people who wrote it, then good for you. You also totally tested it yourself, you didn't just jump on an edgy opinion to feel smart when parrotting it. You totally didn't, nah.

>If the username is already invalid
Complete bullshit poettering made up. Even if he genuinely believed that the username was invalid, the correct thing to do would be not to run the service not run it as fucking root.

Poettering is full of shit like he always is.
POSIX describes exactly how you handle it to avoid this exact problem:
gnu.org/software/coreutils/manual/coreutils.html#Disambiguating-names-and-IDs

...

I'd argue that if the system tools allow the creation of a user, that user should be valid, and systemd has to bend over and take it. Why should systemd decide what user has a valid name or not? It's only concern should be to launch the service the way it should be launched, not do a half assed job out of it because reasons...