You are not logged in.
On another forum, I made the comment that systemd violates the UNIX principle of "do one thing and do it well." Someone responded to that by saying "so does the Linux kernel." I'm not a programmer or coder or anything like that, but I have a feeling that that's an apples to oranges comparison. Am I right or wrong?
Offline
The Linux kernel is an example of a monolithic kernel, which can be contrasted with the microkernel paradigm.
And just to note that "GNU" stands for "GNU's Not Unix" — Linux has never really adhered to UNIX principles anyway. I would also mention that systemd is actually very modular in it's design with *many* compilation and configuration options that can be used to filter out unwanted functionality and also that systemd offers a unified collection of userspace control programs that is *very* similar to the way that all of the true UNIX® systems work but I wouldn't want to start an argument
Brianna Ghey — Rest In Power
Offline
So if I'm reading your response right, you're saying the response I got on the other forum is correct?
Offline
Well I would agree with the comment, yes. But I'm no expert so... ¯\_(ツ)_/¯
Brianna Ghey — Rest In Power
Offline
Well I would agree with the comment, yes. But I'm no expert so... ¯\_(ツ)_/¯
Well, you are more of an expert than I am. But, I'd like to hear what others have to say.
Last edited by Ron (2021-02-06 21:34:10)
Offline
imho, I thought it, systemd stepped outside the FOSS agreement(?)/standards because the source code was not readily readable by a human.
pic from 1993, new guitar day.
Offline
the source code was not readily readable by a human
That's off topic for this thread but just sounds like complete nonsense. Do you have a citation for that?
The systemd code base is licensed under the GPL[0][1] so it does conform to the FSF guidelines[2][3][4].
Brianna Ghey — Rest In Power
Offline
No, I was thinking of pulse source code, and no I can't prove that. Off topic, sorry about that.
pic from 1993, new guitar day.
Offline
imho, I thought it, systemd stepped outside the FOSS agreement(?)/standards because the source code was not readily readable by a human.
readable perhaps...
You might mean not understandable, the project lacks a serious amount of documentation.
Offline
Ive always found this somewhat of a bad design/security flaw in systemd.
machinectl
* machinectl gained a new verb 'shell' which opens a fresh
shell on the target container or the host. It is similar to
the existing 'login' command of machinectl, but spawns the
shell directly without prompting for username or
password. The pseudo machine '.host' now refers to the local
host and is used by default. Hence, 'machinectl shell' can
be used as replacement for 'su -' which spawns a session as
a fresh systemd unit in a way that is fully isolated from
the originating session.
Its been awhile since i was on a systemd controlled machine, but from memory the user is able to gain su privilege with machinectl shell providing no password? Or has this behavior changed?
Last edited by dice (2021-02-07 14:30:50)
Offline
the project lacks a serious amount of documentation
I don't think so:
$ apropos systemd | wc -l
188
$
The man pages are extensive and comprehensive. See also https://www.freedesktop.org/wiki/Software/systemd/
the user is able to gain su privilege with machinectl
If the user is in the sudo group then machinectl(1) can be used to open a root shell with the user's password, which can also be achieved with sudo(8). If the user isn't in the sudo group then any attempt to open a root shell with machinectl(1) is refused unless the root password is supplied.
Last edited by Head_on_a_Stick (2021-02-07 14:41:51)
Brianna Ghey — Rest In Power
Offline
The Linux kernel is an example of a monolithic kernel, which can be contrasted with the microkernel paradigm.
And just to note that "GNU" stands for "GNU's Not Unix" — Linux has never really adhered to UNIX principles anyway. I would also mention that systemd is actually very modular in it's design with *many* compilation and configuration options that can be used to filter out unwanted functionality and also that systemd offers a unified collection of userspace control programs that is *very* similar to the way that all of the true UNIX® systems work but I wouldn't want to start an argument
http://soggywizards.com/tips/code/os/in … -bad-ugly/
Last I checked, SystemD had sixty-eight separate binaries. So it's not directly monolithic as so many of the ill-informed assert.
Offline
^ I said I didn't want to start an argument
Can we try to stay on topic? What do you think about the comment described in the OP?
Brianna Ghey — Rest In Power
Offline
^ I said I didn't want to start an argument
Can we try to stay on topic? What do you think about the comment described in the OP?
I would liken systemd to svchost.exe to answer your question
Last edited by dice (2021-02-07 14:55:08)
Offline
dice wrote:the project lacks a serious amount of documentation
I don't think so:
$ apropos systemd | wc -l 188 $
The man pages are extensive and comprehensive. See also https://www.freedesktop.org/wiki/Software/systemd/
So are the issues.. https://github.com/systemd/systemd/issu … s%3Aclosed
Last edited by dice (2021-02-07 15:30:57)
Offline
So in what way does the Linux kernel follow the Unix principle of "do one thing and do it well"?
Do you not agree that the kernel is even more monolithic and bloated than systemd?
Brianna Ghey — Rest In Power
Offline
So in what way does the Linux kernel follow the Unix principle of "do one thing and do it well"?
Do you not agree that the kernel is even more monolithic and bloated than systemd?
Well that is as what the op mentions as apples vs oranges. Systemd is not a kernel?
Offline