systemD is coming to a port near you!

Well the Linux kernel guys usually dislike kernel drivers which are not GPL licensed, OpenZFS is CDDL. CDDL is compatible with FreeBSD, but not GPL. Many believe that this was done back then on purpose by Sun, because the Solaris release 10 which introduced ZFS in 2006 also had some other groundbreaking innovations like Dtrace in it, and Solaris back then was open source. What Sun didn't want to see is ZFS and Dtrace popping up in Linux instantly, therefore the CDDL. Dtrace was re-licensed under the GPLv2 in 2019 by Oracle, so some still do hope this might happen with ZFS soon as well.

Their Linux kernel developers mantra is: yes user, you can do that, but then don't pester us with bug reports. Your kernel is tainted.

Also this means that they don't really care about providing the interfaces which ZFS might need, or not. In their eyes it's an external thing, and that's all about it. So the developers have always to play the catchup game when a new Linux kernel version comes out.
 
Even if those innovations were permissively licensed, they would still manage to find a way to fsck things up with a bunch of layer violations. The community has an unhealthy habit of reinventing the wheel poorly also. Btrfs, systemd, pulseaudio, systemtap, epoll, etc. are all examples of that. Meanwhile, ZFS and DTrace became first class citizens on FreeBSD.. with no wasted effort no reinventing crap.
 
I advocate porting systemd to FreeBSD but won't use it. There are still people who prefer systemd, and this port will promote influx of those kind of users. It should not be applied to the base system, but there is no drawbacks in just porting them.
 
but there is no drawbacks in just porting them.
If you pull in systemd[1] into FreeBSD, you would have to deal with one of the worst maintainers ever existed.

And finally, the lamest vendor response award went to Systemd supremo Lennart Poettering for his controversial, and perhaps questionable, handling of the following bugs in everyone's favorite init replacement: 5998, 6225, 6214, 5144, and 6237 that we covered here.

Let alone the technical issues, i find it much more important that the person behind the software is reliable, which is not the case here.

[1] https://www.freedesktop.org/wiki/Software/systemd/
Yes, it is written systemd, not system D or System D, or even SystemD. And it isn't system d either. Why? Because it's a system daemon, and under Unix/Linux those are in lower case, and get suffixed with a lower case d. And since systemd manages the system, it's called systemd. It's that simple. But then again, if all that appears too simple to you, call it (but never spell it!) System Five Hundred since D is the roman numeral for 500 (this also clarifies the relation to System V, right?). The only situation where we find it OK to use an uppercase letter in the name (but don't like it either) is if you start a sentence with systemd. On high holidays you may also spell it sÿstëmd. But then again, Système D is not an acceptable spelling and something completely different (though kinda fitting).
 
If you pull in systemd[1] into FreeBSD, you would have to deal with one of the worst maintainers ever existed.
I am pretty sure the FreeBSD developers would reject systemd alone due to its LGPL license.

Aside that making it runnable on FreeBSD would require heavily work, it would be really more of a fork rewriting large stuff in it. And therefore require a dedicated FreeBSD maintainer, who is not Poettering, who is able to deal with Poettering and willing to play the eternal catchup game then with him.
 
I advocate porting systemd to FreeBSD but won't use it. There are still people who prefer systemd, and this port will promote influx of those kind of users.
Would FreeBSD really benefit from having those kinds of users?

A user that will only use FreeBSD because it resembles Linux somewhat is likely to be more effort than they are worth.
 
Poettering's quoted mailing list posting is rather typical, he has made many such postings over the years regarding manipulating users and devs into adoption.
Let's call manipulation Social Engineering from now on. Social Engineering is a cool buzzword for the same thing, as described on Page 1 of "The Art of Deception: Controlling the Human Element in Security" by Kevin Mitnick. He's a hero to some people, I have his book in pdf format.
You see this "culture" right through Red Hat, certain Linux kernel devs, gnome project, Debian and Ubuntu people, Arch Linux, etc where there is this supreme arrogance and widening gulf between developers (many of whom are on the IBM/Red Hat payroll) and Linux users who are viewed in much the same way as a Microsoft Windows customer.
I saw the same thing happen to PC-BSD when X-systems came into the picture. Their own arrogance was what I used against them, called PC-BSD the Win98 of the UNIX World, left and never let them live it down.

Which, once again cynwulf, brings us back to the culture of GroupThink. Which uses manipulation in their tactics of out-grouping. But their cause is just so getting down and dirty is justified. Yeah, that's the ticket, and an example of how the mind can use rationalization as a means of self-defense.

The biggest single problem with SystemD is that Lennart is an a**h*** (you can figure out what the stars in that word mean). Note that I didn't say that he doesn't understand systems research and architecture, or is a bad software engineer.
No, when you couldn't refute his credentials you sunk to a personal attack on his character. It's a standard tactic. One you're above, disparage the use of and tells me something you can't see.

It has become a good base to create embedded systems (Netflix, Juniper, NetApp, Isilon), and as an alternative for hobbyists.
I'm considered to be a hobbyist but don't see the connection to Netflix, Juniper, NetApp or Isilon using FreeBSD. I don't use any of those services and don't even have a PS4.

I'm seeing more and more of that hobbyist use being driven by near-religious beliefs and paranoia, not by rational engineering tradeoffs. In the discussions here and other places, I see the biggest driver of FreeBSD adoption being irrational things like "I hate Linux" or "I hate Lennart" or "I'm old and don't like change".
You are the most knowledgeable person when it comes to UNIX I have ever met, have such a rich background in close association to the early days of UNIX development and are a wealth of knowledge in the area of your expertise I feel we are lucky to have in our midst.

But your area of expertise is not Psychology or Behaviorism, it's mine. A Professional in my field, qualified as such by experience, with 46 years experience my current level of expertise.

Within the space of one paragraph you raised yourself above the gutter in voicing your opinion of him as an a**h*** to lofty heights, casting aspersion on hobbyist driven by "near-religious beliefs and paranoia", not by "rational engineering" based argument. Statements such as "I hate Linux" or "I hate Lennart" or "I'm old and don't like change" being "irrational things".

Obviously, people don't say these things out loud, but hide them behind lofty statements like "Unix philosophy".

That you couldn't see it in yourself and were oblivious to exhibiting the same behavior you decry, as obvious to me as my ways might seem irrational to a hobbyist in my field.

You once said that if someone you considered to be of questionable character walked into the room while you were eating dinner you'd get up and walk out. Look within, Ralphbzs.
 
What other kernels does systemd support anyway?

Someone less weird would surely have just said "non-Linux kernels".

Microsoft's init system also works on every kernel apart from the "niche ones"... So just NT then. Very odd wording. Can't decide if he was being crafty or if he is just out of his depth in terms of wider knowledge.
It shows his Linux-centric view of the world. Every OS is a kernel + GNUish userland. He didn't (and maybe still doesn't) know that the BSDs are more than a kernel.
 
[1] https://www.freedesktop.org/wiki/Software/systemd/

Yes, it is written systemd, not system D or System D, or even SystemD. And it isn't system d either. Why? Because it's a system daemon, and under Unix/Linux those are in lower case, and get suffixed with a lower case d. And since systemd manages the system, it's called systemd. It's that simple. But then again, if all that appears too simple to you, call it (but never spell it!
Yes well Linux ain't unix, so it's wrong assuming we should use unix naming conventions for a mess like systemD. :rolleyes:
 
Wear your love for it! (A design which Poettering featured on Twitter.) This bug report is just hilarious btw: https://github.com/systemd/systemd/issues/437

Ekai6QhVcAEofNW.jpg
 
Distrowatch is not a reliable source for statistics. Nobody knows how many people are using what, where, when ....
Jessie, the owner of DistroWatch, caught people voting multiple times when he had the thread about whether or not to include or segregate Operating Systems that had a Religious or Political stance.
 
  • Thanks
Reactions: a6h
People can follow the guide, but do beginners prefer modifying manually than having GUI in installation?
It's not the installation process that gives people who use my tiutorial or new people in general the most problems, and I've watched closely since mine went up,

It's tinkering around tweaking this file and that before they have the knowledge gained through experience to fix the problems that can and do arise from that practice. Then they become needlessly frustrated, and it can't be their fault (it worked this way on Linux), so they blame FreeBSD and trudge back to Tux Town, tail dragging, 8oz wings flapping.

I've said many times "If it's not broken don't fix it." But nobody listens.
 
Trihexagonal your second paragraph is key to me. We were all newbies once, I remember when one had to really know your monitor and card specs to get it all working.

Is it easier if a single tool could properly detect all the hardware and install everything the user wants? Easier, yes. Better? Not to me because then I am constrained by what the tool-creator thinks is "Best".
A good script or tutorial that gets someone 85-90% "there" is more useful in the long run. Regardless of the OS, a user may need to get out of their comfort zone to fix things (Windows, bad update, regedit are my personal examples); if the needed tweaking can be minimized and "how to do it" is clearly documented (even to the point of absurdity), with maybe clearly "go to here and ask for help", then maybe the frustration turns into a learning experience.
 
f the needed tweaking can be minimized and "how to do it" is clearly documented (even to the point of absurdity), with maybe clearly "go to here and ask for help", then maybe the frustration turns into a learning experience.
I broke it down into exculpating detail for people who were just like me when I was a noob using PC-BSD in 2005 so hopefully they wouldn't have to struggle as much as I did teaching myself. I still hadn't built a desktop from scratch tilI got here and used a tutorial written by someone else to do it.

I'd already taught myself to use ports but was still using their product to get to the desktop. If I'd have continued to hang out there and let them do the hard work for me, I wouldn't be the person I am today. Or what's left that make up the person I am today.
 
  • Like
Reactions: mer
Teach a man to fish, give a man a fish, complain that the fish isn't caviar.

circling back to SyStEmDee, some of the links that someone posted with Potterings quotes/myths, I went and read. The funniest was "systemd is not monolithic because it has 67 separate binaries".
So, ok, technically, he has a point, but if "all 67 binaries are used to usurp major parts of the OS and only work with and for each other" isn't that monolithic in practical terms?
 
I can't speak to all of what motivates the FreeBSD core team, nor do I make a study of it, however, not having such things as a pre-configured desktop environment with extensive prepackaged graphics capabilities, in the kernel and elsewhere, makes base system FreeBSD very compact, and the ideal choice for headless servers and embedded systems which have no eye-candy or graphics requirements.

Furthermore, it does so in such a way as to not inhibit other end-users who might wish to add such things after the initial install. I tend to label this as flexibility. Anyone is free to configure FreeBSD in whichever way one chooses.

Anyone who wants a desktop installer is free to write one of their own, and FreeBSD provides everything one might need in order to do so.

If one also wants automatic graphics hardware detection and configuration, it might turn out to be a pretty rough ride, and require a lot of frequent, ongoing maintenance and OEM-specific customization, since hardware manufacturers are many in number, and continually adding new graphics processors, each with its own increasingly complicated set of OEM-specific requirements.

Adding such capabilities and overhead for graphics hardware requirements, which are constantly changing, and most of which will go unused in any single specific hardware system, might very reasonably be characterized as unnecessary, inefficient, and inelegant bloatware.
 
however, not having such things as a pre-configured desktop environment with extensive prepackaged graphics capabilities, in the kernel and elsewhere, makes base system FreeBSD very compact, and the ideal choice for headless servers and embedded systems which have no eye-candy or graphics requirements.

Furthermore, it does so in such a way as to not inhibit other end-users who might wish to add such things after the initial install. I tend to label this as flexibility.
That's the way we like it and the people who actually use FreeBSD want it to stay.

Anyone is free to configure FreeBSD in whichever way one chooses.
And no two desktops look the same. You don't have to like mine and yours doesn't have to look like mine.

Unlike a wallpaper show/screenshot thread of Linux desktops that all look the same except for the wallpaper.

I'm working on making FreeBSD more a viable option to the Linux community, and for some BSD as a stated alternative to 🌟SysD🌟.

The New World Order Top Secret Illuminati issued code name for the new OS once it assimilates SysV and World Domination put in place as Priestly peyote prophesy preordained previously in the past, probably.
Resistance is futile.
 
I saw this related article in this weeks DistroWatch News. On my Kali box with SystemD running, which I'm using now while ports compile on two other FreeBSD machina:

"People who are interested in porting software between open source platforms such as Linux and the members of the BSD family just gained a new tool: a cgroup filesystem for the BSDs."


It includes a quote from the InitWare/CGrpFS Github page:

"CGrpFS is a tiny implementation of the GNU/Linux CGroup filesystem for BSD platforms. It takes the form of a FUSE filesystem and implements robust tracking of processes. Resource control, however, is not present; the different BSD platforms each provide different mechanisms for this, none of which are easily adapted to CGroups semantics. The process tracking alone is sufficient for the main user of CGrpFS, InitWare, a service manager derived from systemd.


CGrpFS is available under the Modified BSD Licence. It is not as-yet very well tested, but seems to work fine for InitWare's purposes."

 
For those who think FreeBSD needs to change anything: You're wrong.

I just wrote my first init-script, for a software I wrote myself and need myself. I made it "comfortable", so it translates some rc.conf(5) variables into command line options automatically. The script looks like this: https://github.com/Zirias/zfbsd-ports/blob/local/net/remusock/files/remusock.in

So, apart from this one shell function providing the configuration comfort, it's all declarative and automated by the awesome mewburn rc framework, see rc.subr(8). Plus I need to set correct permissions on the directory where the pid file will be created – damn easy to do in a shell script.

I assume this could be solved with systemd as well. Do you really want to know/learn how? I don't.
 
I don't know exactly which problems it solves that are not addressed by other init systems. What I do know is since it was rolled out boot hangs and slow start-up times are very common, which could be forgiven if the rest of it were intuitive and stable while booted. Not so in my experience at all.

Also looking at the responses by the maintainer above about CVEs is very alarming, they do not appear to take security very seriously, which is not good when it is a huge monolithic daemon...
 
Plus I need to set correct permissions on the directory where the pid file will be created – damn easy to do in a shell script.

I assume this could be solved with systemd as well. Do you really want to know/learn how? I don't.
You are so not into modern programming. Systemd defines a whole language for creating pid files and setting the right permissions on them. You really should get with the times. It's make you so much more agile and functional.

Yes, that's sarcasm.
 
Back
Top