So does anyone here actually give a fuck about systemd beyond

If so, why?

I guess not.

>>62214195
Still not production ready, it does not respect sysctl ulimits and requires setting them up manually in a service file.
Writing init files for systemd in trickier than for sysvinit.

It's cool to rant about sustemd because of "muh unix philosophy" or "muh feature creep" but for the people who actually work with init scripts and the like, systemd is a godsend

>>62214195
>muh boot times
Enjoy your botnet.

>>62214195
I'm kinda torn on it - on the one hand, its annoying as shit having to relearn a whole slew of things that I was already comfortable with for what seems like absolutely no tangible benefit at all. From my point of view I've yet to find something that wasn't made more awkward or just broken out the box On the other hand having everything homogeneous 'should' make administering things easier.

Bottom line, If I want to call myself an administrator I have to be able to administer everything that's put in front of me, and that means systemd so I deal with it and learn.

Like it or not, while it may not quite be the only game in town, it's a big player regardless of my opinion.

Poettering and his crew are a disaster. They introduce bugs in systemd that makes it incompatible with other software when used in a specific way and force others to write a workaround in their software. They ignore serious vulnerabilities because they think everything else handles this in the same faulty way, which is not true. On top of that, Poettering accuses the FOSS community of being generally shit, while they're only acting this way towards him because they're fed up with the bullshit he produces.

Red Hat is defending Lennart and his shit, too. They managed to force major distributions to use systemd and PulseAudio by default. Stuff like GNOME, which doesn't need to interact with the init system suddenly depends on systemd.

systemd's codebase is huge and rapidly changing, which makes auditing it next to impossible. Add that to "Team Poetter"'s general incompetence, their bug-ignoring behavior, the drama they create around bugs and the huge attack vector created by a single huge binary running as the one process that brings the whole system down if it crashes. At this point it's hard to say why anyone deems systemd secure.

>>62214195
It hasn't pissed me off yet. Its also made it very easy to set network triggers when my laptop wakes up from sleep. I don't like that Firefox is going pulseaudio-only though

>>62214195
Does anyone have that image of the smelly guy who dislikes systemd that uses devuan distro and has a foil hat?

>>62214529
Has this been proven yet? I understand the hate for it becoming too canon, too monolithic, etc. but to jump from that to -> malicious software seems irrational. pls explain && no bully

>>62214195
I thought wikileaks proved its an NSA backdoor

>>62214857
auditing your moms pussy is next to impossible

>>62214195
Yea, I like systemd. It's fast, pretty easy, pretty sensible, logging works okay by default...

It's a nice init. And it made things easier vs needing to have busybox and a dozen other things in some hard to diagnose initrd. Everything required being included in systemd is nice.

>>62215079
>(You) think people hate systemd for no reason other than
>>systemd sux xd
>provide reasons
>ur mum xd
kys

>>62215114
snort yourself out, bucko!

https://www.youtube.com/watch?v=n_4UFMp19tE

>>62214195
Stoped using it about 3 months ago, never looked back.

It's a pretty good OS but lacks a good init

>>62214195
what happens when systemd is sold to a parent company that happens to be oracle?

>>62214529
>>62214984
I've asked for MONTHS on /g/ for anyone to point to the lines of code in systemd that is a "botnet" and to date, no one as been able to do this.

>>62215054
Is this suppose to be a joke?

>>62214857
So would openRC be a better alternative?

>>62214195

Its a huge security hole and I wish it wasnt combining a bunch of shit into a huge lump that will absolutely end up with a back door.

We are inevitably going to get hit with a system d crypto virus at some point.

>>62216207
This is grade A propaganda. I wish the dwindling anti-systemd fags would just kill themselves already.

>>62216346
>300k lines of C code? Sure! Just fuck my shit up, poettering!
t. (You)

>>62214507
It's a godsend for people WRITING init scripts. It's a disaster for people USING init scripts.

>>62214195

>use open source systems in part to leave behind the mess that was MS
>open source systems eventually move to implement a similar mess

I'm checking up every now and then to monitor the progress of systemd free systems. inb4 write your own. People that are comfortable with systemd are 12 year olds who have always have foss systems around and take then for granted. It's a horrible move.

>>62215276
It gets forked

>>62214857
Harry Pottering and your moms pussy.

>>62216375
So the code is entirely open source and available to anyone to inspect, but your complaint is you're too fucking lazy to do that? Seriously, go kill yourself.

>>62214195
Call me when it is secure and doesn't just hand out root provides to any user that wants it.

>>62214195
I am honestly curious what the advantages to systemd are.

Why should a distress use it?

>gnome now depends on systemd

THEY SAID IT WOULD JUST BE A INIT SYSTEM

REEEEEEEE

>>62217845
Unironically install Gentoo
https://wiki.gentoo.org/wiki/GNOME/GNOME_Without_systemd

>>62217845
No, it is basicly the next gnu. We will have to call linux
>gnu/systemd/linux

File: 1483860895271.gif (1MB, 430x360px) Image search: [Google]

1MB, 430x360px

>>62214195
I don't even know what init systems are

>>62218577
The kernel starts init, and init starts all the other programs.

>>62214344
>Writing init files for systemd in trickier than for sysvinit.
>in trickier

Maybe for you.
They are both easier and more capable.

>>50%ofposters
/g/ is actually kind of entertaining once you come to the realization it's not your job to educate the willfully ignorant.

>>62216565
*closed*
*wontfix*
>CVEs don't exist®
Open source is meaningless if the main developers don't care about legitimate and we'll documented flaws

>>62214857
This^
The current status of major Linux distros is just sad.

Even worse there isn't any other option.

OpenBSD maybe? Fuck this, and fuck redhat shills

>>62218797
>/g/ is actually kind of entertaining once you come to the realization it's not your job to educate the willfully ignorant.

t. arrogant guy

>>62214195
Yep, it's much more resource-intensive than sysvinit, poorly documented, under constant revision and expansion, has had countless security issues including multiple arbitrary code execution and privilege escalations, and the developer is a complete piece of shit who works for multiple intelligence agencies to subvert Free Software.

It's why I am staying with old Debian on my one computer, and moved to OpenBSD on the other.

It's cuckware for transfags.

>>62214857
>Red Hat is defending Lennart and his shit, too.
This, and I'm sick of getting shilled by Red Hat. I will never support systemd or Red Hat or Fedora for this reason.

It's also why I will never buy another Intel CPU. Their shilling is out of control and I want to make sure their business is harmed by their thoughtless actions.

>>62215054
It did, but the CIA niggers are thick on 4chan so they will just flail around wildly when confronted with this.

They ask you to point to existing code in systemd which has vulns, why do that? Much better to sell them to people who really want them, and do more damage to systemd in the meanwhile.

One can point to the countless escalation and arbitrary execution bugs which have been discovered, though, or point to the Google pings it sends, and so on.

>>62215194
Linux is pozzed anyway now. Huge bloated kernel which will never be audited, more and more un-Unixy stuff, SJW companies ranting about whites, it's a shit show.

>>62216565
>So the code is entirely open source and available to anyone to inspect
Freeze it for a year. I dare you. If you do people will peek into it and discover innumerable back doors.

>>62217843
It boots very slightly faster than sysvinit, and provides a number of agencies with handy back doors into your system.

It's a win win.

>>62219368
OpenBSD, 9front, TempleOS. These are the new /g/ operating systems.

>>62214195

The advantages it claims to provide over sysV were solved better by other init systems without the problem of subsuming a million other purposes with newer and buggier implementations. It's basically not possible to run without it on distributions where it is default. Everyone knows it only caught on because it had Red Hat behind it.

>using loonix

just install win7

>>62219631
Red Hat is cancer.

>>62220588
Confirmed pozzed by CIA.

>>62220621
>CIA
like CIAfags have extra time just to spy on your sorry little anime ass.

>>62220660
They will spy on anybody, data storage is cheap.

However when they send shills to /g/ they aren't sending their best, as you prove.

>>62216392
Can you imagine the poor IT folks who have to type systemctl restart apache instead of /etc/init.d/apache2 restart; pkil apache2; sleep 10; pkill -9 apache2; /etc/init.d/apache2 start?

The people using the init scripts must hate systemd

>>62222956
>not making an alias
also
>init makes it possible to actually solve problems even if systemd breaks

>>62219631
init would still be the default on Debian if they hadn't killed Ian.

I've been using systemd with arch and it's been a soft ride until now. I don't know why the hate really.

>>62223347
tons of security holes, nearly 1/2 million lines of code, inability to understand such a complex system if it breaks, constant rapid revision

>>62222956
>rc-service apache2 restart
wew lad

>>62223347
I wouldn't call my feelings towards it "hate", more like mild dislike.

It replaced a bunch of single-function tools that did their jobs well, with one tool that tries to do everything, but not as well.

I've had a bunch of production machines at work that I've had to reboot via sysctl from the console, after systemd stopped responding to signals.

I've run into some hisenbugs after upgrading systems to distros that include systemd where it doesn't quite behave in the ways the docs say it should, some processes don't get restarted despite what the configs say, etc.

It's got a huge number of dynamically linked libraries (run ldd against it) when I still feel that init should be statically linked, and allocate all of the memory it needs when it first starts up. No dynamic loading, or dynamic memory allocation in process 1; it can't rely on any system resources being available at any time past startup.

It feels like a step backward for no advantages.

>>62223445
>tons of security holes
so many you couldn't even list a single one

>inability to understand such a complex system if it breaks
translation, I'm too dumb to understand it

>nearly 1/2 million lines of code
translation, I'm too lazy to examine it

isn't it backdoored?

>>62223675
https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2017-1000082

Systemd would be great if it did what the docs said. Alas they were written as goals, and not a description of how the system works.

Personally, I feel that a traditional init, and deamontools would meet 90% of what people claim to need from systemd.

Some issues I've run into:
* Random refusal to restart processes after they exit, with nothing in the logs
* The main systemd process locking up, and neither restarting processes, or allowing the system to reboot
* Refusal to boot with "unusual" USB drives attached (in my case, multiple format card reader.with a dozen or so different card slots)
* Bug reports being confronted with Pottering claiming the reporter is an idiot, and they should just use an undocumented workaround (the --force --force --force --force --force option to systemct for example) instead
* Random logging failures where it decides to stop listening on the log socket; since it has to impose itself in front of any logging daemon that is running
* Required reboots to do upgrades; upgrading a process in place while keeping the same PID has been a solved problem for years (fork a new worker, hand off current work to the new worker, exec the new instance with an option that tells it to reclaim the current work from the external worker).

Overall it seems to be badly designed, and buggy as a result. It doesn't do what it claims to do, but when you mention that fact, the maintainers go nuts.

>>62223675
>so many you couldn't even list a single one
There's a whole list of the ones that have been discovered.

How about root access for any user name beginning with a number?

Just Lennart things.

>>62223894
I love how the fucking thing hangs for five fucking minutes if you don't have your ethernet cable plugged in on boot.

>>62214195
systemd is a good idea

systemd actually manages to bring common interface across Linux distros for service management, power management, session management, etc. This is shit is pretty important, since it reduces amount of useless work maintainers and developers have to do to maintain compatibility across distros.

Although there are things I don't like about systemd and I'll avoid it until the whole thing is going to get ironed out.

It's monolithic. It just feels counter-intuitive. I like Gentoo's initiative to split code out of systemd (elogind, eudev) keeping API compatibility.

It still can get magical. Debian unit for a example: https://anonscm.debian.org/git/collab-maint/openvpn.git/tree/debian/openvpn.service

Also Poettering. Poettering thinks it's okay to keep things broken just to prove his point. See writeable efi mount issue for example, or flat volumes in pulseaudio. This is how 6 years old behave.

File: worried laughter.png (514KB, 520x678px) Image search: [Google]

514KB, 520x678px

>>62223675
>https://cve.mitre.org/cgi-bin/cvekey.cgi?keyword=openrc
>https://cve.mitre.org/cgi-bin/cvekey.cgi?keyword=sysvinit
>https://cve.mitre.org/cgi-bin/cvekey.cgi?keyword=systemd

>>62216417
FreeBSD 11.1 is really nice. No systemd. Native up to date ZFS that works as the root filesystem. Linux binary compatibility.

>>62223896
Don't forget closing the bug because defaulting to root is a feature.

File: typical.png (4KB, 407x123px) Image search: [Google]

4KB, 407x123px

>>62224131
I'm closing this post.

>>62224142
kek

File: SmugSJW.jpg (5KB, 275x183px) Image search: [Google]

5KB, 275x183px

>>62224131
>>62224142
>a bug
>no goyim, not a bug
lol

File: 1455337734660.gif (496KB, 500x455px) Image search: [Google]

496KB, 500x455px

>>62224142
>typical.jpg
BAAAAAHAHAHAHHAHAHAHA

>>62223894
Actually, systemd has a reexec option for updating, but since PID 1, enjoy your kernel panic if the reexec fails.

>>62224215
Wait; they can't handle an exec failing?

Rather than properly handle the error, they just let things panic?

That's even worse than not handling the case in the first place.

>>62214857
Newfag here, can you recommend a Linux distro without it? I was interested in the Sourcemage threads when they were up.

>>62215114
Good reason faggot

File: 1504042305843.jpg (59KB, 645x412px) Image search: [Google]

59KB, 645x412px

>>62224314
Yep but this is 'expected' behavior in case there are any bugs in any systemd's config files.

What you don't like your OS crashing if there is a typo? The expected behavior is your machine hangs, dummy. Not that your single service start/restart should fail.

You see, by redefining the expectations they can make a lot of people swallow the biggest cocks.

>>62224353
FreeBSD. It's not Linux, but it will probably work for you.

If you need something which is available only as a binary, it can run Linux binaries, if you install the right libraries.

>>62224385
Lennart's the one you're replying to, and like every SJW he's been trained to 'deplatform' anything he disagrees with.

He will simply slide around any legitimate criticism, it's his way.

He's from communist East Germany after all.

>>62224404
I already know about BSD but isn't there a Linux distro without this crazy new subsystem?

>>62217828
This shit again, please learn something about the topic before you post. Or tell me, how exactly do you get root just by creating user with special characters. Prerequisites

>>62224449
make a service file with User=0day
run the service
it escalates the service running as 0day to root privileges.

>>62224426
You could always try Slackware.

>>62219368
And fuck you, go jerk of to your waifu

>>62219404
Show us the way, oh master

>>62219512
Sure buddy supa haxxor

>>62224410
You talk like sjw, and btw you still didn't say the reason except sjw, hurry durrr he's a next Hitler

>>62224468
But to make a service you need to be root

>>62217845
This. systemd is slowly taking over everything. Well if it acts like a botnet and if it looks like a botnet, then...

>>62224628
Not an argument. Systemd is wrecking Linux, which is overbloated in any event.

>drop support for a lot of old hardware in the name of compactness
>codebase still grows at exponential rate
>change ABI so people can't use old hardware
>add systemd botnet
welcome to windows town

>>62224677
this was possible with daemontools like a decade ago

>>62224688
at least now we might have muh gaymes on loonex :(

File: Screen Shot 2017-09-02 at 6.12.07 PM.png (27KB, 614x150px) Image search: [Google]

27KB, 614x150px

>deleeted his idiotic comment
This was possible with daemontools a decade ago.

>>62224711
Over half the entire Steam library works on Linux, every pre-SJW-era game works on Wine, literally what shit-tier games are you even playing which won't work with Steam or Wine?

>>62224720
daemontools isnt avalible on linux :^)
daemontools isnt systemd :^)

>>62224732
witcher 3

>>62224780
>daemontools isnt avalible on linux :^)
yes it is
>daemontools isnt systemd :^)
that's the point

>>62224790
has worked fine for months

>>62224780
>daemontools isnt avalible on linux :^)

WTF are you talking about?

https://cr.yp.to/daemontools.html

>>62224900
You see now how systemd defenders are willing to actually just baldly lie about their botnet to cause its spread?

It's really a sad state of affairs.

>>62224900
mount :^)

>>62224938
does that carrot smiley indicate that you know you're lying and being obtuse, but it's insurance so you can call everything a joke later?

you should be killed in a just world, or your tongue torn out

Still not a single valid argument why systemd is bad.
>its monolithic
It consists of individual fucking binaries
>its insecure
yeah, it also has a backdoor for CIA niggers.
That bug with numbers in usernames is funny though.

>>62225023
>Still not a single valid argument why systemd is bad.
>>62223894
>>62223554
>>62219631
>>62219404
>>62214857

You might want to actually read the thread before posting.

>>62225023
>bugs
>backdoors
>huge and complicated
>not unixy
>insinuates itself into every facet of OS
>sponsored by spy agencies
yes arguments ahve been presented, ignoring them is making you look like an actual retard

File: zhen1.jpg (217KB, 749x321px) Image search: [Google]

217KB, 749x321px

SYSTEMD KILLS

>>62223445
>inability to understand such a complex system if it breaks
The people saying this should look at the Linux source code sometime.

If you hate using complex system, you should stay away from computers entirely.

>>62224468
>escalates
That word doesn't mean what you think it means, entry level IT scum.

>systemd shills will defend 3+ minute boot times

>>62225390
>The people saying this should look at the Linux source code sometime.
All 14 quadrillion lines?

Everybody who cares about security moved away eons ago.

>>62225417
Unplug your ethernet cable, it adds 5 minutes to your boot time.

>>62214195
No.

>>62223717
quite a few backdoors have been discovered thus far, yes

I tried a systemd distro (Ubuntu) but kept getting weird errors, ended up putting OS X back onto my mac instead.