SYSTEMD ABSORBS L2TP

L2TP has been officially absorbed by SYSTEMD.
github.com/systemd/systemd/commit/54c258566ea1e6cf0c0c51d8a1afb8cdc0c92379

How many years until all VPN code is officially merged, making SYSTEMD 100% exploitable?

Attached: Screenshot_2019-03-14_12-09-49.png (999x532, 78.26K)

Other urls found in this thread:

github.com/systemd/systemd/blob/5176b272ea459e1753f007e0534dba5eceec42b5/test/fuzz/fuzz-netdev-parser/25-wireguard.netdev
github.com/systemd/systemd/tree/54c258566ea1e6cf0c0c51d8a1afb8cdc0c92379/src/network/netdev
git.zx2c4.com/WireGuard/tree/
wiki.archlinux.org/index.php/WireGuard#Using_native_systemd_support
github.com/systemd/systemd/issues/11992
github.com/systemd/systemd/issues/11949
lists.freedesktop.org/archives/systemd-devel/2010-September/000391.html
archive.is/3i2d6
remote-lab.net/linux-l2tp-ethernet-pseudowires
github.com/systemd/systemd/issues/5043
twitter.com/SFWRedditVideos

Attached: wireguard.png (500x545 49.83 KB, 262.88K)

I hunger!

also absorbed

btw

prove it, faggot

github.com/systemd/systemd/blob/5176b272ea459e1753f007e0534dba5eceec42b5/test/fuzz/fuzz-netdev-parser/25-wireguard.netdev

it literally replaces the wireguard binaries with a pure systemd implementation

I bet the UNIX hater will somehow blame C and the Unix philosophy for this. He's already claimed that systemd's binary logs are a good idea that we weenies simply don't understand.

Attached: 1470872437387.png (564x676, 7.77K)

github.com/systemd/systemd/tree/54c258566ea1e6cf0c0c51d8a1afb8cdc0c92379/src/network/netdev
here it's beside l2tp

systemd is much more than an init system. it's more akin to something like a userland part of kernel. wireguard integrated into muh linux kernel? what a shame. why the userland kernel part can't provide integration code?

Learn English

What is Poettering's endgame?

are you one of those anons that are still under an influence of that old canonical anti-systemd psyop?

from what I can tell, this is code to support wireguard within systemd-networkd, sorta like how OpenVPN has to be supported inside networkmanager, connection manager, Wicd, etc. It's not actually replacing Wireguard, as you'll still need the stuff at
git.zx2c4.com/WireGuard/tree/
for it to function, or you could just run this alone without the systemd shite and move on with your day.
Wireguard needs to support Slackware, Gentoo, Alpine, and FreeBSD too, so they're not going to tie themselves to Systemd, especially since their stated goal is a small attack surface and low LoC.

That's a funny response to someone telling a shitter to learn English. Are you alright?

Attached: 1464731760456.jpg (1080x1080, 736.69K)

according to:
wiki.archlinux.org/index.php/WireGuard#Using_native_systemd_support


i don't think their intention is to "eat" wireguard or l2tp, but probably the support code (as the connection code is integrated into kernel) is simple enough to be fully integrated

I decided to check the systemd github because my blood pressure was too low and out of morbid curiosity

github.com/systemd/systemd/issues/11992
< poettering commented 13 hours ago

github.com/systemd/systemd/issues/11949
< poettering commented 3 days ago

kys

How the fuck didn't you fags who say to hate systemdicks haven't explored this yet? SYSTEMD IS OPPRESSIVE AND NEEDS A THICK NON-GENDERED COCK YESTERDAY

What is L2TP and how does it affect openvpn?

Slowly but surely, covering the whole kernel until it can be the kernel.
I know some fucking weed in my garden is doing the same, covering other plants and stealing all the sun, bottom plant dies and become fertilizer.

It's an older protocol and it doesn't affect openvpn.

fucking capped

Attached: system d capped.png (1832x599, 291.71K)

why????

Attached: 5721a7f95d93551687dfb74868a25680137f286b2e1a8f2424b38cbe761a8bbc.png (600x450, 125.43K)

Well user, microsoft windows has its SSL and vpn libraries in its init too. So the CIA/redhat/pottering is just replicating what they see inside the windows source code. It makes perfect sense when viewed from the angle of desroying linux and making it windows.

lists.freedesktop.org/archives/systemd-devel/2010-September/000391.html
archive.is/3i2d6


not worth it and cross-distro unification is worth more.

The cancer grows.

it sure is if you want that it replaces windows. compatibility is important for some users and now you have to maintain multiple packages so it works on all distros or even different versions of the same distro so most people will stay on the proprietary systems that just work for them

But why? Why is systemd like some fat fuck at a family buffet, taking every little fucking thing they can and piling it onto their plate even if it doesn't make sense? In the end, it's just a slop of shit.

Attached: but feathers.png (458x458, 134.08K)

looks like systemd is some giant shadow project to centralize entirety of internals for linux for Red Hat and soon IBM

There's a easy way around this. Remove any libraries that require GLIB and glibC. GLIB, the gnome library, speaks for itself. Glibc is a bloat peice of shit with huge amounts of uncessary backwards compatibility and muslc or uclibc are better replacements. Once you do those two things your system will by defintion be pottering free as glib is a neccessary requirement for dbus, policykit, systemd, gtk+3, consolekit, modern udev, and a few others that will go unmentioned.

Actually glib, and gdbus which is different then dbus by extention being removed from your system means you can not have any pottering software installed as glib is a neccessary common library for all pottering software, and I do mean all because gdbus is needed for compatibility with older zeromq botnet shit.

Someone should submit a pull request that includes all of X11.

This but doesn't he censor everything negative on the systemd github?

Match made in heaven. Systemd finally managed to absorb something worse than itself.

What's wrong with l2tp?

systemd offical wayland compositor soon.

Root problem: L2TP is emulation of dialup connection, rather than just tunnelling (like GRE, IPIP or Wireguard).

1) L2TP does not tunnel IP, it tunnels PPP. Extra layer of complexity without any benefit, except compatibility with old enterprise setups.

2) Tunnel & channel establishment. For protocol that encapsulates unordered messages into unordered messages it makes no sense, but is there to match modem pool usage patterns. Compare with IP-in-IP, that lacks control messages or state at all; or wireguard, which has only key-establishing handshake and optional heartbeats.

Also no encryption. It's might be good from purity point of view, but in practice unencrypted tunnel is mostly useless today, so in any practical case you also deal with IPsec, which is not prettiest thing too.

Attached: l2tp.png (548x140, 10.33K)

I don't see what the problem is with PPP then. I never programmed with it, so I don't know what the raw protocol is like, but when I used it decades ago it worked fine.
Isn't the whole point of it being Layer 2 Tunneling Protocol that it tunnels the network layer? Is it because it's low-level instead of application layer like openvpn or whatever else?

Wow, something truly shittier then systemd. Never though i'd see the day.

Wow, something truly shittier then systemd. Never thought i'd see the day.

There is nothing wrong with PPP itself, it's just that it is not necessary for IP tunnelling case.

Network level is 3 (ip). Level 2 is data link (ethernet, ppp).
Yes, it is (assuming s/network/link/).

Application level is, surprisingly, specific application protocols. HTTP, SMTP, etc. OpenVPN's tunnelling is network level (but it can also work as level 2 (ethernet) tunnel).
Yes, because it's too low level, in sense that is has to emulate details that are irrelevant for common usage. Basically, you need L3TP in majority of cases - but L2TP is widely used for historic reasons.

L2TP is non-properierary, but supported by equipment manufacturers; works over UDP (as opposed to stuff like GRE, that works over bare IP packets); works together with IPsec terminating equipment and ppp access/billing infrastructure. For ISPs and business such factors are more important than protocol cleanness and simplicity.
But for personal cases, without burden of legacy and need to integrate with RADIUS server that serviced dial up customers for 15 years, there is no reason to use L2TP.

Thanks. I appreciate the info.

There is L2TPv3 which is completely different to L2TPv2 and does what you expect a protocol named L2TP to do.
It's a simple static Ethernet over IP pseudo-wire adding tunnel and session ID headers with optional UDP header.
remote-lab.net/linux-l2tp-ethernet-pseudowires


IPsec has that design-by-comity stank with every possible option because Cisco and the NSA couldn't agree on a single back door.

om nom nom nom nom

No you stupid shitskin. When you ask a question, you don't just stick the interrogative word at the beginning of a declarative sentence in order to convert it into a question. You have to invert the sentence as well.


When you fail to do this, you sound like a kebab vendor and you instantly incur the wrath of every native english speaker who reads your scrawl. So stop fucking doing it.

Get this,
Systemd is trying to be that next monolithic thing. It's only a matter of time until it absorbs linux itself and tries to remove the legacy code we had for decades.

It'll be the next windows 2.0 where it's either you upgrade all your hardware or stay in shit or forces you to stay in shitty architectures only which also prevents the competition for alternative architectures (win-win).


Not Poettering. It's CIA. wait until they threaten linus*


This. Based and red-pilled.

Also L2TP is supposed to only work in the firmware level - a router that uses a hardened busybox and non-systemd kernel. If you put L2TP in the software, it'll be easymode for the cianigger to fuck yer shit up.

Weren't they able to already own your system by just receiving malformed packets from someone?
Also

This
Didn't it make sense to you that systemd is a landing cushion for windows?
From init states in GUI, registry-like confs, timers that imitate win task scheduler, and many more like crashdumps, event log/viewer, and its binary logs. It'll grow up one day and you'll realize how much of a shitload it is considering the fact most device, web servers run linux and so does smartphones too. It'll never go away. It's CIA.

They already neutered Linus with the CoC. Disposing of him won't be something that happens overnight. It will be one of those things where people look back and go
And Linus will give everyone some bullshit excuse like he's moving on with his life.

yes.
the code does pose restrictions on how the l2tp layer is implemented. therefore systemd is reducing your spectrum of viable options.

Attached: th.jpeg (217x287, 21.09K)

the /genius is wrong of course.
he is so wrong that i won't even bother to indulge in futility.
i like to enjoy his genetic retardation.
he smells like a nigger too.

Attached: 1433878772353.jpg (1162x850, 280.47K)

pic related. it's cancer.

Attached: index.png (211x238, 5.27K)

putting a power drill through your hard drive would be faster.

github.com/systemd/systemd/issues/5043
Reminder that until next release, systemd lacks functional interface bonding.

HAPAS ARE SUPERIOR TO WHITES

HAPAS ARE SUPERIOR TO WHITES

HAPAS ARE SUPERIOR TO WHITES

HAPAS ARE SUPERIOR TO WHITES

HAPAS ARE SUPERIOR TO WHITES

Masons, Masons everywhere.

Europeons need to burn in a tar-pit. Disgusting

Masons, Masons everywhere.

I smell rats.

Yeah, fuck off boomer. Nobody cares.