Hello, I'm somebody who wrote a low level linux utility and I want to release it under a free license. However, I have inspected systemd and got terrified by the style their developers treat the community. Therefore I want to ensure my software never gets maintained or picked up by systemd developers. Which license should I chose to accomplish this?
Thank you for your help.
What you want is by definition not a free license.
If you were using Debian you could simply uninstall systemd (assuming that you weren't one of the 22 people using gummiboot (soon to be called sd-uefi)).
Seriously -- there are no reliable reports of systemd doing what these anonymous trolls report it as doing.
One of the advances of systemd is that it does log stderr from processes it starts, so some clown has decided to complain that it doesn't.
There are no reliable reports of this behaviour -- it's all a bunch of AC posts on Slashdot plus one on Reddit (which was immediatley refuted), plus a few paranoid claims about bug reports being "deleted". How do you delete a bug report from bugzilla? You can't.
No, it's the kernel. systemd is a crap load of applications. Applications that ignore stderr, drop higher priority syslog messages, and ignores nonzero exit statuses.
What the fuck is up with you trolls repeating these stupid lies?
systemd may or may not have problems, but it demonstrably does not "ignore stderr, drop higher priority syslog messages, [or] ignore nonzero exit statuses".
If it does, or ever has done, any of these things where is the fucking bug report.
Firstly, Multics, while good, was only different from Unix in degree, not kind. Multics also included a panic routine and would call it when it was unable to continue.
HODIE NATUS EST RADICI FRATER ring any bells? (Just one example).
Secondly, have you been using any Multics systems lately? I wonder why not.
Thirdly, which meaning of "security" are you using today? You seem to be chopping and changing.
Several solutions rely on setting up a proxy with a whitelist of sites, or similar, but I'm afraid this approach can make internet unusable, or otherwise be easy to bypass
> run on sysvinit then the exit code is also ignored.
Wow, you systemd guys lie a lot. It is a serious bug, and your example is contrived. Of course if you intentionally ignore the exit status by appending a & at the end of the line, then the exit status will correctly be ignored.
Except that with systemd the exit status is not ignored.
The contrived "systemd_broken" bug was exactly the same thing as my broken init script -- launch a process that fails after being forked and then complain that systemctl start didn't report the error.
But systemd status does report the error.
Try it for fucks sake.
# cat broken_systemd.sh #!/bin/sh echo "Oh noes, the daemon is dead!" >&1 exit 1
# cat/etc/systemd/system/broken_systemd.service [Unit] Description=Broken systemd example After=network.target
[Install] WantedBy=multi-user.target # systemctl start broken_systemd # echo $? 0 # systemctl status -l broken_systemd.service * broken_systemd.service - Broken systemd example
Loaded: loaded (/etc/systemd/system/broken_systemd.service; disabled)
Active: failed (Result: exit-code) since Thu 2015-02-19 12:31:13 CET; 19s ago
Process: 4586 ExecStart=/root/broken_systemd.sh (code=exited, status=1/FAILURE)
Main PID: 4586 (code=exited, status=1/FAILURE)
Feb 19 12:31:13 celtic broken_systemd.sh[4586]: Oh noes, the daemon is dead! Feb 19 12:31:13 celtic systemd[1]: broken_systemd.service: main process exited, code=exited, status=1/FAILURE Feb 19 12:31:13 celtic systemd[1]: Unit broken_systemd.service entered failed state.
not "supposedly" - *really*. if you run apt-rdepends -r libsystemd0 | a bit of awk | sort | uniq there are four *THOUSAND* five hundred packages that, if you were ever to do "apt-get --purge remove libsystemd0" you would NEVER be able to install. that's a whopping FIFTEEN PERCENT of the entire debian package repository that you are prevented and prohibited from installing, should you ever make the decision that you wish to keep libsystemd0 off of machines that you manage!
And why do you want to keep libsystemd0 off your system?
Android does not use linux's/sbin/init, or sysvinit, or OpenRC or upstart or systemd. It has its own init system which does not run shell scripts (Android doesn't even have a shell by default AFAIK).
(By the way, why all the bash love? Only an idiot would write init scripts in bash. Anyone worried about security would use a POSIX shell like dash).
This systemd pdf article is pretty unremarkable except for what is written in big font in the 2nd page:
Another aspect of systemd is that it collects all output from processes that it starts.
Since systemd launches all processes, it can easily spy on all the process outputs and transmit that to whichever TLA it wants. This is a major spying attack.
systemd puts the stderr (and sometimes stdout) of processes it starts in the system journal. This is extremely useful (to the extent that various anti-systemd trolls around here are denying that it happens).
If your journal is readable by the NSA you have bigger problems than systemd. If your various init started daemons are writing confidential information to stderr you have bigger problems than systemd.
systemd does not "launch all processes". systemd only has access to the stderr/stdout of the processes it does launch.
Slashdot Mirror
Yeah, I forgot the damned "</b>". Makes me sound like an angry person. Not that I could ever be mistaken for an angry person otherwise.
Maybe you should look harder. "systemd-sysv-" means "REMOVE systemd-sysv".
The systemd-shim is to make pam-systemd work without systemd being installed. gdm3 needs pam-systemd.
If you don't need/want gdm3 just remove pam-systemd and don'n install systemd-shim.
Hello, I'm somebody who wrote a low level linux utility and I want to release it under a free license. However, I have inspected systemd and got terrified by the style their developers treat the community. Therefore I want to ensure my software never gets maintained or picked up by systemd developers. Which license should I chose to accomplish this?
Thank you for your help.
What you want is by definition not a free license.
Take the "do no evil" license from JSON.
Yes, exactly. I'm running Debian Jessie and I'm not really comfortable with binary logs.
The default configuration on Debian Jessie is to log everything to syslogd as before.
What difference did you notice in the logging?
I'll be forwarding my logs to nice text files for the foreseeable future though
Why are you doing that? The system already does it for you.
If you were using Debian you could simply uninstall systemd (assuming that you weren't one of the 22 people using gummiboot (soon to be called sd-uefi)).
Seriously?
The GP posts a straight out lie, some moron mods it "interesting".
I point out that it is a lie. I get moderated "flamebait".
Even if you hate systemd, try to beat it with facts. If you start to base your arguments on verifiable lies we begin to doubt your sanity.
Even if those "technical arguments" are lies?
Seriously -- there are no reliable reports of systemd doing what these anonymous trolls report it as doing.
One of the advances of systemd is that it does log stderr from processes it starts, so some clown has decided to complain that it doesn't.
There are no reliable reports of this behaviour -- it's all a bunch of AC posts on Slashdot plus one on Reddit (which was immediatley refuted), plus a few paranoid claims about bug reports being "deleted". How do you delete a bug report from bugzilla? You can't.
No, it's the kernel. systemd is a crap load of applications. Applications that ignore stderr, drop higher priority syslog messages, and ignores nonzero exit statuses.
What the fuck is up with you trolls repeating these stupid lies?
systemd may or may not have problems, but it demonstrably does not "ignore stderr, drop higher priority syslog messages, [or] ignore nonzero exit statuses".
If it does, or ever has done, any of these things where is the fucking bug report.
Ow! That hurt.
Firstly, Multics, while good, was only different from Unix in degree, not kind. Multics also included a panic routine and would call it when it was unable to continue.
HODIE NATUS EST RADICI FRATER ring any bells? (Just one example).
Secondly, have you been using any Multics systems lately? I wonder why not.
Thirdly, which meaning of "security" are you using today? You seem to be chopping and changing.
Several solutions rely on setting up a proxy with a whitelist of sites, or similar, but I'm afraid this approach can make internet unusable, or otherwise be easy to bypass
Exactly like all the "solutions" on Windows.
You've never heard of Willy Soon?
http://en.wikipedia.org/wiki/Soon_and_Baliunas_controversy
The "deal with science" bit is (among others) "On Past Temperatures and Anomalous Late-20th Century Warmth", Eos, Volume 84, No. 27, 8 July 2003.
This is another example of "if we can't discredit the science, discredit the scientist for being paid to do research."
This is Willy Soon we're talking about here. The "science" was discredited long ago.
Which says nothing about security.
The 'original UNIX' was a mess compared to what?
Before multics nobody had a fucking clue about security.
> run on sysvinit then the exit code is also ignored.
Wow, you systemd guys lie a lot. It is a serious bug, and your example is contrived. Of course if you intentionally ignore the exit status by appending a & at the end of the line, then the exit status will correctly be ignored.
Except that with systemd the exit status is not ignored.
The contrived "systemd_broken" bug was exactly the same thing as my broken init script -- launch a process that fails after being forked and then complain that systemctl start didn't report the error.
But systemd status does report the error.
Try it for fucks sake.
Oh, yes -- the only mention of "cancer" I can find on the systemd mailing lists is from anonymous at dodgeit.com who says:
Kay, Please go die in a fire along with Lennart. Your type is the cancer that
is killing any semblence of usability Linux once had.
I doubt Kay is your mother.
> Have you reported a bug?
Why? So you people can tell me again that you hope my mother dies of cancer? I think I've filed nine bugs in the past two years we've used systemd
So, which is it? Have you filed a bug or not? If you have, where is it? What was the resolution?
One configuration file in a non-turing complete language is not "a mess of scripts".
If you read it it's an event driven system, closer to systemd than sysvinit.
not "supposedly" - *really*. if you run apt-rdepends -r libsystemd0 | a bit of awk | sort | uniq there are four *THOUSAND* five hundred packages that, if you were ever to do "apt-get --purge remove libsystemd0" you would NEVER be able to install. that's a whopping FIFTEEN PERCENT of the entire debian package repository that you are prevented and prohibited from installing, should you ever make the decision that you wish to keep libsystemd0 off of machines that you manage!
And why do you want to keep libsystemd0 off your system?
Do you know what libsystemd0 does?
(Hint: it does nothing if systemd is not pid 1).
You want me to stop, while you're being the fanboy. Why don't you stop riding Poettering's dick?
What the fuck happened to you, man? Your ass used to be beautiful.
In fact, even on Android the init system is a hobbled together "mess of bash scripts"
You think that's a bash script?
You might like to check out https://android.googlesource.com/platform/system/core/+/master/init/readme.txt
Android does not use linux's /sbin/init, or sysvinit, or OpenRC or upstart or systemd. It has its own init system which does not run shell scripts (Android doesn't even have a shell by default AFAIK).
(By the way, why all the bash love? Only an idiot would write init scripts in bash. Anyone worried about security would use a POSIX shell like dash).
I may be wrong, but I thought OpenRC was a replacement for sysvinit, not "initd" (whatever that is).
(I.E. OpenRC replaces the horror that is /etc/rc?.d but still uses /sbin/init).
This systemd pdf article is pretty unremarkable except for what is written in big font in the 2nd page:
Since systemd launches all processes, it can easily spy on all the process outputs and transmit that to whichever TLA it wants. This is a major spying attack.
systemd puts the stderr (and sometimes stdout) of processes it starts in the system journal. This is extremely useful (to the extent that various anti-systemd trolls around here are denying that it happens).
If your journal is readable by the NSA you have bigger problems than systemd. If your various init started daemons are writing confidential information to stderr you have bigger problems than systemd.
systemd does not "launch all processes". systemd only has access to the stderr/stdout of the processes it does launch.
I did you damn liar. I've opened several, and they've all been ignored.
No you did not you lying troll. If you had made a bug report you could provide a link to it so we could see what actually happened.
You ignore and delete bug reports.
Ah, here it is, the paranoid claim.
"you delete bug reports".
This is simply not possible in any bug reporting system in use today (without hacking the database by hand).
WE NEED stderr FOR TROUBELSHOOTING!
And the notable feature of systemd/journald is that syserr is saved in the journal, not quickly scrolled off the top of the screen.