Why the continued separation from Linux?

Sorry if this is offensive, but why the continued separation from Linux? I know that they both started separate, but I feel like both projects could get a lot of good from a merging. The BSD software could get a lot more eyes upon it, and Linux could gain more ingrained support for things like HAMMER and ZFS. Sorry if this is a rudimentary system, but this question has bee bugging me, as the only solution I could think of is the license difference, but that doesn't seem a very big snag.
 
Catholics and Lutherans originated from the same place and could gain a larger following and help each other, too, but that's not going to happen due to philosophical reasons.

The biggest concern, outside licensing, is that FreeBSD is a complete system while Linux is only the kernel. The core of an OS. Every Linux distro is run by a totally separate, disparate organization. FreeBSD is one organization. The same is true for openBSD and netBSD.

And never the 'twain shall meet.
 
Implying a separation just because there are different ways is at best inconsiderately.
From the same point of view I could ask you why do you separate from your neighbour, because of not having the same furniture, driving the same car, wearing the same clothes, cheer for the same team... - having the same live-style, while certainly there are points you could improve by doing it like your neighbour.
You could also ask why Linux is not becoming more BSD.

Doing things differently is not only some kind of freedom but the most important part of evolution.

Both worlds are not separated that strictly.
Besides there are not a few developers working in both worlds, at least the whole universe of OpenSource is overlapping, which makes most of what the user eventually needs for its daily usage.
To put it in a nutshell:
If most of what you do on a computer is e.g. using firefox, LibreOffice, and VLC, it doesn't really matter much, which system is below.
Several Linux-distros offer fully automatic installation and configuration of everything.
Downside: You have to stick with what others have chosen.
If you don't like that, such as me and many others, but want your own personal, individually tailored system - one does not need to create "a new distro for everybody" again, then you need something modular, like FreeBSD.
Downside: Your effort is needed. Nobody else can read your mind, and deliver you your system as you want to be it exactly - except youself.

And besides that, both worlds look and learn from each other.
Not only the things could be done better, but at least important also to avoid the mistakes, the others did.
It's not up to FreeBSD if Linux implements ZFS (which they are doing.)

And in anyway anything has to fit the concept.
In my eyes that's a core difference of BSD and Linux - the concept of how things have to be done.

In my eyes in Linux-universe there is either the total open, less restrictive, more easy, loose way to do things, a creative playground, which produces masses of software-packages - but also lots of junk and useless crap,
or there are those projects under the lead of some "dictatorhip" - those produce the functioning turn-key-distris; which came a long way to get there, baby - a long way.
Trying to install some Red-Hat-Linux in the late 90s from several diskettes, getting an X-server up and running on those tube-monitors, fumbling hours only with mode-lines on a flickering, distorted screen (if any), and such adventures, are war stories nobody today believes anymore who doesn't made those experiences himself.
And those who did for sure don't want to get back there again.

I fumbled a lot with many Linux many years.
All this "sudo apt get...", "sudo apt foo...", "sudo apt bar..." (just to mention one of today's Linux-installation methods), just for getting "Error... error... error... Aborted!" (almost Everytime, with any Linux-Installation tool) really pissed me.

To me an OS is not for to end in itself, install it, tinker with updates, handling errors, being happy when finally something simply runs, and waiting impatiently for the new version to come just to have again something to tinker with.
I want/need to use the software.
System's administration is a necessary must, but not the purpose for a computer, which by the way is not just a personal opinion.

The first time on FreeBSD I did a pkg install ... or a make install ... within ports tree, and the shit simply ran from the start until the very end without any complains, no aborting errors, simply produced an then installed, and working piece of software at the very first go - and it worked reliably everytime.
That's what sold me FreeBSD.

This for me is for certain a crucial point I definitely do not want "more Linux" in FreeBSD.
Absolutely not! For sure not in this point. No discussions, not compromises. No!
It's not "old-school old-farts are not capable of changing anymore", it's simply the experience not to get back into the moisty trenches, again.

Linux-universe is full of people having lots of fancy ideas.
But that's not the first things one needs.
Even if many are not agreeing with me, not willing it to understand, especially not the ones with fancy ideas,
but the crucial point of the famous journey over a thousand miles that starts with the famous first step is it will not reach a destination if the thousand miles are not traveled until its end.

It doesn't matter if it's an OS, an airplane, or a simple craft kit: The most important thing is, the job needs to be done in full. Otherwise you end up with some partially-finished mess "to be continued someday", stored away in the attic. Linux-universe has a large attic of such things.
The more comprehensive, and the more complex a system becomes, the more people are needed.
The more people are involved, the more organization becomes vital.
Something many do not accept. Especially the "creative ones" sense it as an offensive, ideas-killing-environment, feeling restricted in their liberty, and start looking elsewhere for "more freedom", for more understanding for their ingenious ideas, running from pillar to post, instead to learn that work most of the time is tedious, and simply has to be done. And not by others, only.

To get something reliable one need reliable people, finishing tedious work reliably on time.
For that organisation is needed. Way - way - more than fancy ideas.
Which means less liberty, but more efficient usage of resources, and reliable success.
It's the very nature, the whole point of organization, to define what has to be done, and what not.
Of course, one may discuss how organization is to be done - before you start.
But then one has to decide for and stick to it, and if so do the job. Or beat it, before.
Discussions about organization while a running project over and over again is mostly just another way to simulate effort, while bringing in new, fancy ideas again, instead of doing real work, while using up resources needed for doing the work.
Again, not old-fart's paranoia but simply real-life experience.

When the work is done, and resources are left, then one may start to realize new ideas.
Anything else will sooner or later only end up in the attic.

I could think of is the license difference, but that doesn't seem a very big snag.

If you are a private user, doing only private things for personal use, then you don't need to bother about if the software you're using is under GPL, MIT, Apache, or BSD license.
But if you want to spread your work to others, especially to be used by others, or even do something professional, particulary want to earn money with it, then you better take a very good and close look at licenses, because especially the GPL contains dangerous traps you may naivly run into, may shut your project down, cost you money, or both.

Besides that you cannot simple discriminate "Linux = GPL, FreeBSD = BSD license" - which is true for the very system (Linux: the kernel) - but that's not all a system consists of.

The makers decide under which license they publish their software.
So you'll find all kind of OpenSource lincenses in both worlds.
Examples:
Gimp is under GPL; doesn't matter if you run it under Linux, or BSD.
So is gcc.
clang is Apache, no matter if it's under BSD or Linux.

So if you're not want to do anything for personal, private use, only, anymore, you better take a good look at all the software that is involved in your process, at each of its licenses, and potentielly contact the license owner, or change to some other piece of software.
 
why the continued separation from Linux?
why not?
BSD was around for years when Linux was starting out. so you could also ask why it separated itself from BSD / Unix.


no, thanks.
another edit. by all means, go ahead and do it. I will stay where I am comfortable and try it out if, when it looks promising. but pupil already made some good points why it's unlikely. besides that, mixing everything together is one of those linuxy ideas I don't understand, and honestly, I don't want to. simple as that. just accept it, please.
 
OpenZFS is an excellent example of coordination. From release notes for FreeBSD 13.0 (April 2021): "… The ZFS implementation is now provided by OpenZFS. …".

The base system – FreeBSD – is, technically, not something that can be merged.
 
Even if you were only going to compare Linux (=only a kernel) with the FreeBSD kernel—which I encourage you to do, even in a very global way and from a fairly distant point of view—you'll notice that philosophical differences, (and perhaps to a lesser extent: organizational** & licencing differences) have led to a different take on kernel structures and therefore implementation*. I dare to state that neither party has any significant intention or incentive to change towards the other and would be willing to agree on some common path. And that is just fine in a varied landscape of multiple OS-es.

It has been suggested that if the licensing troubles of (Free)BSD in its early days hadn’t happened, Linus might have been using FreeBSD and developing for it, instead of developing a new kernel on his own. That’s all water under the bridge now.

In the light of all the differences between these two different *NIX-es (i.e. (Free)BSD versus Linux distributions) I think it is remarkable that (Open)ZFS has managed to acquire such a large gathering and has succeeded in uniting many Linux users & users of other OS-es in their use and support of ZFS; and in uniting OpenZFS developers from both worlds in support of ZFS for a multitude of OS-es.

___
* One aspect of such a comparison would be the scheduler. If you are interested in that, have a look at The Battle of the Schedulers: FreeBSD ULE vs. Linux CFS by Justinien Bouron, Sebastien Chevalley, Baptiste Lepers, and Willy Zwaenepoel - USENIX ATC 2018, as mentioned in McKusick’s An Overview of Scheduling in the FreeBSD Kernel

** Edit: this to reflect organizational aspects on the level of actual persons; not organizational aspects on the source code level.
 
Last edited:
...I could think of is the license difference, but that doesn't seem a very big snag.
This is a much bigger deal than you think, especially with the move to GPL-3. Yes, I know the Linux kernel will never go to version 3, but the impetus behind a lot of BSD-licensed software that has emerged over the past decade or so comes from private companies who do not want to depend on GPL-3 software.
 
Homogeneity is bad. This should be especially obvious given the events of the past few days.

This is true with almost everything in life - you shouldn't plant just one crop. You shouldn't rely on just one supplier for critical supplies. You should have multiple ways of generating power. Etc etc etc. As far as life lessonns go, this is right up there with keeping things as simple as reasonably possible and planning for, or at least, being aware of, the worst case scenario.
 
  • Like
Reactions: mro
"Why isn't BSD more like Linux?" "Why doesn't BSD embrace systemd?"

"Hey Siri, what is a 5 letter work starting with T ending with L that often lives under a bridge and you have to pay a toll to?"

Besides everything everyone else has said, Linux camp is often not welcoming of "not invented here". Go back and search the kernel mailing lists about ZFS for a simple example.
 
Diversity is a good way to ensure robustness, no matter if you are talking about operating systems or other things.
Yes. Actual point in case - crowdstrike.
This is where German railway had a good day, they still use windows 3.11 and were not impacted by this. My work also was not interrupted, I have a debian system at $job. At home, I run FreeBSD.
 
why the continued separation from Linux?
There isn't really. A number of projects to merge the two have existed:

Debian kFreeBSD - FreeBSD kernel, Linux userland
Chimera Linux - Linux kernel, BSD-ish userland
MkLinux - Linux as a subsystem in Mach, similar to the BSD subsystem

I also feel that at the kernel level, OpenZFS probably pulls a lot of BSD code into Linux. Likewise the BSD graphics stack is pretty much grabbed from Linux.
 
There isn't really. A number of projects to merge the two have existed:

Debian kFreeBSD - FreeBSD kernel, Linux userland
Chimera Linux - Linux kernel, BSD-ish userland
MkLinux - Linux as a subsystem in Mach, similar to the BSD subsystem

I also feel that at the kernel level, OpenZFS probably pulls a lot of BSD code into Linux. Likewise the BSD graphics stack is pretty much grabbed from Linux.
I tried Gentoo/Freebsd more than a decade ago. It was ok, but at the end of the day you wind up with something that is neither Linux nor Freebsd. "Ni chicha ni chocolate" like we say in Spanish.
 
Linux developers sec are by definition kernel (and associated kernel modules I imagine) developers. In FreeBSD you have an all-inclusive team for the base, closely linked to all those who contribute to the ports tree. There is of course a difference in scale between FreeBSD and the large teams of various Linux distributions. However, after I have started my FreeBSD journey, I have come to appreciate the ports tree (be it ports or packages) more and realize its importance for the FreeBSD community; including my respect for all those who contribute.

Besides everything everyone else has said, Linux camp is often not welcoming of "not invented here".
Yes*. I think that is also a sign of different kind of developer communities; perhaps one could even characterize that as a cultural differences.

The thing that cannot be underestimated is cultural differences in different organizations, probably even more impactful when most or largely consisting of non-paid volunteers without an economical/financial relationship. Developer communities are the most directly tied to the development of the product they support, that extends into user communities. In my view there are big differences. Limit that for instance to just the BSDs: among the three BSD communities there is a kinship like relationship and I like to think users and developers alike appreciate and respects each other’s efforts and products. But even given that, there is no desire to integrate netBSD, OpenBSD and FreeBSD, and in my view, that's perfectly understandable and to be respected. Each excel in their own way and evolve according to their own vision.

___
* ZFS, (Porting the Solaris ZFS file system to the FreeBSD operating system by Pawel Jakub dawidek & Marshall Kirk McKusick) seems a nice example of a FreeBSD community that is willing and able to incorporate something external of a substantial size and complexity.
 
Sorry if this is offensive, but why the continued separation from Linux? I know that they both started separate, but I feel like both projects could get a lot of good from a merging. The BSD software could get a lot more eyes upon it, and Linux could gain more ingrained support for things like HAMMER and ZFS. Sorry if this is a rudimentary system, but this question has bee bugging me, as the only solution I could think of is the license difference, but that doesn't seem a very big snag.
I for one value diversity as a general principle.

Imagine you nominate your favourite dish and the next day all others but this one would be abolished. Would you like a world like that? I don't. Or take music. Which is the best song? You'll have it 24h/7. And all should have it alike. A terribly totalitarian idea btw.
 
I for one value diversity as a general principle.

Imagine you nominate your favourite dish and the next day all others but this one would be abolished. Would you like a world like that? I don't. Or take music. Which is the best song? You'll have it 24h/7. And all should have it alike. A terribly totalitarian idea btw.
Having options is good, but some people call it "fragmentation" or whatever other stupid word.
I'm from the opinion of "you don't like it? don't use it. Use what suits you better".
 
still use windows 3.11 and were not impacted
So, the lessons learned by this CrowdStrike desaster are:
1. Don't use "security-software". It'll cause more damage than malware.
2. Don't upgrade your system. The older the version the less the chance malware attacks you.
?
Before I get quartered up: It was a joke!

I bet at this moment there are CEOs banning CrowdStrike from their systems,
which is right (such things must not happen; this is the price for greed when saving costs for whatever the cost),
but also prohibiting any security-software making this point.

Another site came up on HN the last days:
The Six Dumbest Ideas in Computer Security

Most interesting to me is how this misconception leads to the arm-race the good guys cannot win,
which could be avoided at the start by anticipating basic principles.
The autor sees it as an engineer.
A salesman sees it completely converse.
To me it clearly fits into Microsoft's overall strategy.
Producing cludge not only saves money but also sells more cludge supposedly tinker garbage, not really fixing it.
If this is done by additional vendors - and as far as I remember [until Windows 7] MS does not deliver Firewall/Antivirus-Software automatically with their systems (especially not claiming their systems are designed such are unneccessary/useless) but annoys the user to additionally buy some if one hasn't installed any, recommending certain ones - this also brings them backing support from "independent" third partys, which of course will not tell their customers:"MS is crap. Try something else."
A lot many small companys lives by "repair-services" for Windows:
cleaning up the registry, defragment the disk, install antivirus,...
Also many of the ones doing this to polish up their reputation at family, friends, and neighbours want to stick to MS for the same reason.

That's also why continue to separate from Microsoft. (To justify this post in this thread.)
 
%pupil -sceptical The point, also in this thread, is why w3.11 is not affected. Back in the day there were more viri for windows than you could shake a stick at. Even the old DOS junk still worked, it was one hell of a monoculture. So they made their infrastructure isolated from the rest. The win3.11 boxes only talk to the back ends, possibly some old SUN servers or whatnot. There is a separation, what can infect the client can't do the server. These systems do not need this crowdstrike nonsense, so they don't have it. There is no AD in sight, which removes another attack surface. These systems were designed to be protected and isolated. Today is a different matter. Everything has to be reached from everywhere, which includes the bad actors. This can't work much longer.
 
Back
Top