Hey Sup Forums,
Aside from systemd not truly adhering to Unix philosophy, is there any other good reason to oppose it? systemd .service files are easy to use, supported on every major distro, and I've had no issues whatsoever the last couple of years I used systemd.
Other urls found in this thread:
suckless.org
mirrors.dotsrc.org
github.com
linux.com
artixlinux.org
linux-magazine.com
linux-magazine.com
linux-magazine.com
linux-magazine.com
linuxjournal.com
hackernoon.com
digitalocean.com
digitalocean.com
infoworld.com
blog.darknedgy.net
github.com
github.com
plus.google.com
freedesktop.org
launchpad.net
media.ccc.de
twitter.com
And the larger conspiracy is systemd being part of NSA operation ORCHESTRA
mirrors.dotsrc.org
Use OpenRC or runit. If you use Arch and don't wanna switch just install it from the AUR and do a crond job to remove all .service files from /usr/lib/systemd/system/* daily.
Isn't orchestra partly about people overblowing minor issues distracting from real problems, just like the people complaining about systemd do?
Apart from cancer cells not adhering the philosohpy of your own metabolism, is there any other good reason to oppose them?
"vendor" lock-in
cancer-like behavior
if you depend on one component, you depend on the entire shebang
not user-replaceable
if one component has a bug (or "lack of feature", e.g. devs only have SSDs, so fuck features that make HDD use cases workable) you get bent
before systemd, you could just swap with another component, which maybe even had devs without their heads up their asses
in short, shoved down our throats, takes away our freedom ("linux is not about freedom" poettering fags expected), makes the system more brittle with more SPoFs, devs are not known to prioritize quality, robustness and reasonable backward compatibility (compare to the Linux kernel)
more features, more game-breaking bugs, less choices to unfuck yourself when the devs are behaving like systemd devs
welcome to your brand new RedHat/Windows world, hope you like needing more support contracts to reduce the risks from their overcomplicated shoddy engineering monstrosity
Whole point of Linux is audited software or you're a million times better off with a closed source os. Systemd won't be any time soon and it's only getting bigger
>systemd Breached One Million Lines Of Code In 2017
>Written by Michael Larabel in systemd on 2 January 2018
Fuck off, NSA agent.
1) if something wrong with your fstab os won't boot
2) when it runs longer it eats more resources
3) logging system is broken
Systemd won't be what? Audited, of course it is, don't be silly.
How to audit this behemoth..?
Go here: github.com
Holy shit I thought it was only 200,000
>Audited, of course it is, don't be silly.
pffffft
good one
to be fair it includes dbus, probably udev and other shit, but it's the same shitware from the same team of engineering cowboys
so I'd still count it toward the 'do not trust' LoCs
fun how 'do not trust' suddenly became PID 1, run as root
>more than 800 contributors
>implying it doesn't have a rigorous system for auditing contributions
Stop pretending to be ignorant, it isn't particularly funny.
>tfw you have to run a daily cronjob to remove SystemD shitware
I thought gnu/Linux was all about choice.
I don't see what's wrong with having virus in my computer.
I meant that it's too large to audit single handedly.
Daily reminder that systemd is free software.
It does have issues from a purely technical pint of view, but the botnet accusations are ridiculous.
Any public papers about it?
If you're going to go through the code line by line, yes. But so is any open source project of a considerable size.
Don't forget to tip your tinfoil. :^)
Do I look like fucking google to you?
The project itself has plentiful of documentation, if that's what you're asking for.
Also, Linux.com has a bunch of articles about it, so does the Linux Journal. As for academic papers, you can google for that yourself.
We need an anti-redhat general on Sup Forums.
Fuck gnome, systemD, Pulseaudio
>Aside from systemd not truly adhering to Unix philosophy
Linux is not Unix
wew lad
Not to mention that "the unix philosophy" is a meaningless phrase used as a catch all when someone runs out of technical arguments.
*tips tinfoil for a fellow fag/g/ot* :^)
It's not standardized in any way.
It can't be since it's feature set and interfaces are constantly being changed.
It's very opaque and a pain to debug.
And it boots in non-deterministic order, which can cause unexpected failures to happen on the same previously working system.
Linux is a Unix, it's just not UNIX(TM).
What a disappointment. So you pulling stuff out of your ass? If systemd was audited by a third party (ideally) and not by RedHat or systemd devs, then there surely be at least an article about it.
Fuck off with your idiocy. If no proof then keep your shitty speculations to yourself.
All four statements are wrong.
Everybody notices there is correlation with extreme feminism across tech circles and wacky engineering.
How easy is it to use OpenRC on Parabola? Is OpenRC compatible with .service files or something?
I am writing this message from the glorious Artix linux: artixlinux.org
Has OpenRC by default, Artix is probably the best distro from not-cucked distros !
>Wrong. Here's a random unrelated link
?
>Arch based
into to .Trash with you
>So you pulling stuff out of your ass?
I'm not, no.
>If systemd was audited by a third party (ideally) and not by RedHat or systemd devs, then there surely be at least an article about it.
There are a bunch of them, what the fuck are you on about?
linux-magazine.com
linux-magazine.com
linux-magazine.com
linux-magazine.com
linuxjournal.com
hackernoon.com
digitalocean.com
digitalocean.com
linux.com
infoworld.com
>hurr durr no articles about systemd
>random link
>pretending to be retarded
It explains the architecture of systemd, from the standard features and interfaces to boot order.
it's not like you provided a mountain of evidence for your post either buddy
Why do keep posting random usage articles?
Not one of them are about audits.
It only gives you a shallow overview that's not related to anything I said.
This article explains in depth how systemd resolves the boot order: blog.darknedgy.net
>systemd
>redhat
>industry
>standard
Hey NEETs, how's it going!
There's no any reasons to oppose it, it's a good tool for pajeet butthurt. Sup Forums is pajeet nowadays if you didn't know.
>Not one of them are about audits.
That's not what you requested, fucker. You're moving the goalpost because you are a retard that doesn't understand open source development. The process of contributing is explained in the project's own documentation, it's not like they hire auditors, developers themselves do the auditing.
>It only gives you a shallow overview that's not related to anything I said.
It directly contradicts your claims. It may be a shallow overview, but it's a lot more than you provided.
There's also the Arch wikis and the Debian wikis.
>blog.darknedgy.net
I'm trusting the Linux Journal and Linux.com over a blog called "dark 'n' edgy" any day.
this
Are you fucking retarded? I'm not talking about documentation, but: some competent at security team, audits systemd for possible backdoors or other malicious code/behaviour.
Don't throw me fucking tutorials about systemd you moron.
>If systemd was audited by a third party and not by RedHat
github.com
Most of the people listed here have no affiliation to Red Hat.
>or systemd devs
That's not how the open source development works.
But it's not a standard at all.
SysV init is actually standardized.
That was different user. And I specifically asked for audits, not docs. See:
>That's not what you requested, fucker.
Not the same guy. But it seems pretty clear to me that he meant articles about auditing.
>It directly contradicts your claims
How so?
>not truly adhering to Unix philosophy
That's a good thing, though.
They don't have any. It's a schoolkids' nihilism like the botnet meme and other shit on the neo-Sup Forums.
Not an argument. RedHat is a leader in it's own right. Autists can define however many standards they want, the industry leader will pave the way.
Moving the goalpost, are we?
>some competent at security team, audits systemd for possible backdoors or other malicious code/behaviour.
Nobody does this, not even the Linux kernel is "audited" in this way. Systemd has several contributions from people with an infosec background, though, but apparently that isn't good enough for you. For example Michael Biebl, which is also one of the top contributors to NetworkManager (and responsible for loads of crypto stuff).
Also, several contributors are also kernel devs. For example Till Kamppeter.
>How so?
>not a standard
It's clearly an industry standard.
>feature set and interfaces are constantly changed
They're not, they're pretty much unchanged since 2015 which is as far back as I'm bothered to go in response to your obvious shit post.
>it's very opaque and a pain to debug
This is obviously a subjective criteria, but most devs find it easier to work with systemd due to the powerful control utilities.
>And it boots in non-deterministic order
Which is an outright falsehood, since dependencies are explicitly expressed in systemd.
So? It's still not part of any standard, not even LSB.
RPM, is also from Red Hat and actually standardized.
Should we all switch to RPM now?
Linux is Unix-like you mongol, but not real Unix
>ITT: People who can't use:
>systemctl status *INSERT SERVICE NAME*
>journalctl -xe
HaHa :^)
But I'm not moving my goal posts. It always was and still is: is systemd audited by competent team? If so, show me proof.
I see that you do not understand where the problem is.
- no secret that RedHat is working with NSA
- systemd comes mostly from redhat
- systemd very aggressively pushed as standard
How can you trust anything coming from redhat?
>It's clearly an industry standard.
Look up what a standard is.
And how does your link disagree?
>They're not, they're pretty much unchanged since 2015
github.com
And how does your link disagree?
>This is obviously a subjective criteria, but most devs find it easier to work with systemd due to the powerful control utilities.
Even kernel devs find it impossible to debug the simplest things:
plus.google.com
And how does your link disagree?
>Which is an outright falsehood, since dependencies are explicitly expressed in systemd.
You're an idiot. The dependency graph may be well-defined. That doesn't mean the traversal is. It's literally in its man page:
freedesktop.org
And how does your link disagree?
>is systemd audited by competent team?
Yes, the devs comprise a competent team.
>inb4 hurr durr third party company
That's not how this works. Even the Linux kernel isn't somehow audited and verified by a third party company.
>- systemd comes mostly from redhat
According to the list of contributors, that's clearly untrue. Among the top ten contributors listed in the following link, only three are Red Hat employees.
Hell, even the head of Linux Mint, top devs in Cannonical and two kernel devs are listed here.
launchpad.net
>- systemd very aggressively pushed as standard
While true, this is proof of absolutely nothing except initrd being dogshit.
>How can you trust anything coming from redhat?
This isn't a "Red Hat project", that's not how open source development works. Most contributors have zero affiliation with Red Hat.
I'm done with your trolling.
>Yes, the devs comprise a competent team.
>thinks the devs are doing actual audits
>thinks there's nothing wrong with the people developing the software also auditing it.
Wow, you're so fucking clueless about software development it's not even funny.
Do yourself a favour and stop talking.
>>thinks there's nothing wrong with the people developing the software also auditing it.
Explain to me how you believe the Linux kernel is being developed, this ought to be funny.
Auditing and developing are different things.
Just look up what an audit is before embarrassing yourself further.
>- no secret that RedHat is working with NSA
>- systemd comes mostly from redhat
All the major developers of the Linux networking stack (net-next) are Google employees or former Google employees. Dave Miller, which is the lead, is a former Google employee and currently employed by Red Hat. All auditing is done by these devs and ultimately merged by Dave Miller before he sends a pull request directly to Linus Torvalds.
80-90% of the contributions to networking stuff in the kernel are from Google employees.
It is no secret that Google is working with the NSA, and even a botnet themselves. Can you really trust networking in Linux?
>Contributing to code is same as auditing it
You might know part of it, but not all 1M lines
I'm fine with kernel and most open source projects, that doesn't mean that I have to trust them. And especially systemd. It could be just a solution to init, but why they had to make it this big? That's not right. And if NSA is involved, they'd obviously would want to control as much as they can.
You can't prove me wrong.
Auditing is a part of the bazaar development model, you fucking sperg. Devs audit themselves, there is no "third party company" that somehow is willing to review the Linux kernel (for free) and then give it their stamp of approval. Neither would kernel devs be willing to have to wait while auditing is in progress, they would much rather prefer active development.
>but why they had to make it this big?
Because it has to integrate with a bunch of different standards.
>You can't prove me wrong.
I can prove how ridiculous it is to trust Linux but not systemd, see
ITT: People who need to read the Cathedral and the Bazaar
I am unironically curious about the auditing process in FOSS because I have no idea how it works. Typically, software audits are not done by the developers themselves. Some links or information about the auditing process or processes for the Linux kernel, for something like Debian, and maybe something like Firefox would be interesting to me, thanks.
So basically you're saying that there's no public audit of shitemd, which is what people was requesting, and what you said, existed.
How much of a sperg can you be?
>I am unironically curious about the auditing process in FOSS because I have no idea how it works. Typically, software audits are not done by the developers themselves. Some links or information about the auditing process or processes for the Linux kernel, for something like Debian, and maybe something like Firefox would be interesting to me, thanks.
There are essentially two ways, the Cathedral and the Bazaar. The former approach: a dedicated team works on a project internally and release the source for every release. This is how GNU, GCC, clang, Firefox and the major distros etc are developed.
Then there is the bazaar, which is how systemd and Linux is developed, where there everything is public all the way. In this model, contributors send in patches based on the most recent branch to a dedicated mailing list (each "component", if you will, has their own mailing list). Some trusted devs will review your patch and either acknowledge it or reject it. If a patch is ack'd, then it is thoroughly reviewed and tested by the lead devs of that mailing list and then ultimately taken in (or rejected if it breaks anything). Then once enough patches and fixes have been made, Linus Torvalds will notify everyone saying that they're going to make a release candidate. Leads will make pull requests for their respective responsibilities, and Torvalds will then merge in each branch into the mainline.
For each release candidate, vigorous testing is done and if nobody reports anything wrong, it will make it as a full-blown release.
>no public audit
There is, it is on the Internet for everyone to see.
>third-party audit
Not the same as a public audit.
Why do you lie? that's not good.
>Why do you lie?
?
The github link has been posted several times.
For public audits, there are some community projects, e.g. OCAP.
Some people are crazy enough to enjoy going through third-party code, like this guy media.ccc.de
There's no audit on those links though. Why do you keep lying?
Gas yourself Poettering