What are your thoughts on FreeBSD?

Have you ever used FreeBSD or any of it's derivatives? What were your thoughts on it?(USER WAS WARNED FOR THIS POST, You should always elaborate on your opinions rather than just spilling the soykaf.)

>practially linux but with soykaf driver availability, less programs and no docker why would i use this if not out of curiosity or or in some enterprise edge cases?

I've finally decided to ditch Windows and I'm split between Linux (Ubuntu, Debian or Mint) or some derivative of FreeBSD. Going purely on what I've read in the manual, I'd prefer a FreeBSD but sadly there's all sorts of compatibility issues and just less stuff for it.

In late 2010 I switched to FreeBSD specifically for their stable ZFS implementation. As a Gentoo user switching to any Unix-like system was easy for me. I wasn't happy with the hypervisors available for FreeBSD at the time so this also led me to learn iSCSI so I could get the hypervisor I wanted on one machine and the storage I wanted on another. I lead system deployments for some unincorporated private groups (fancy way of saying "our friends"). Exposure to memory corruption vulnerabilities is just not an option for us. https://redox-os.org/ is not yet ready for prime time. Because we are not a commercial entity, https://grsecurity.net will not work for us, we have the money but their model won't work for us. HardenedBSD is the only project that provides something we can rely on. We aren't gaymers, everything we ran on Linux compiles fine on any BSD. https://hardenedbsd.org/content/easy-feature-comparison https://hardenedbsd.org/content/about > HardenedBSD forked the FreeBSD codebase for ease of development.

>>70924 Thank you for providing a constructive answer.

>>70916 Wasn't Linux like this a decade ago as well?

>>70924 We also landed on HardenedBSD because ZFS and something like illumos Zones are a must-have for us. We are looking for a "hardened illumos" and if we don't find that out there we might try to roll it ourselves. We do have the funding to dedicate engineering resources for this.

>>70930 >Wasn't Linux like this a decade ago as well? The BSDs are older than Linux so that's not an excuse.

>>70930 BSD hasn't seen the kind of wide spread adoption as Linux. Considering it's a descendant of Berkeley Research UNIX, its limited to academic and professional environments. The user base is also pretty small, although I've heard some variants of FreeBSD like Ghost and Midnight can run anything Linux can.

>>70944 >BSD hasn't seen the kind of wide spread adoption as Linux. Unix and BSD were massive before the AT&T lawsuit. >Considering it's a descendant of Berkeley Research UNIX, its limited to academic and professional environments It was used plenty in industry. Especially since the license means you can use it in commercial products with no legal complications. >I've heard some variants of FreeBSD like Ghost and Midnight can run anything Linux can. FreeBSD has a Linux compatibility layer but HardenedBSD disables it by default so it's probably not as secure as you might want it to be.

>>70935 What about Oracle Solaris?

>>70948 >FreeBSD has a Linux compatibility layer but HardenedBSD disables it by default so it's probably not as secure as you might want it to be. How does leaving a niche feature disabled introduce a security concern? Wouldn't less code mean less surface area and less that can go wrong?

>>70952 Oracle Solaris is unfortunately closed source, no way to do supply chain security.

>>70909 ZFS is the only reason to use it over OpenBSD, and OpenBSD is more comfy in my opinion.

Is there a BSD with a non-soykaf license yet? How's HyperbolaBSD coming along these days?

>>70968 As long as it's free/libre software, the choice of free software license shouldn't matter that much, unless you want to become a contributor. >https://www.openbsd.org/policy.html I think making a new thread on /r/ about licenses, copyright, intellectual property and so forth might be better than discussing it here. /r/ is better than other board for this kind of discussion imo. >How's HyperbolaBSD coming along these days? I think it's dead/stalled?

>>70958 The post is saying that the feature is disabled by HardenedBSD therefor the feature is probably not secure.

>>70972 HyperbolaBDS is not dead since it's still being worked on. They just posted an article about it a couple of months ago alongside a roadmap for the project. https://www.hyperbola.info/news/hyperbolabsd-is-an-operating-system-and-not-a-system-distribution/

>>70967 I use freebsd with UFS tho

>>70967 What exactly makes OpenBSD more "comfy" compared to FreeBSD?

>>71008 coreutils more minimal, security is a priority on openbsd, for example they were the first to implement aslr into their os in 2003. freebsd only implemented this around 2019 or 2020

>>70935 Please, if possible, do! I've been waiting for a decent alternative to OpenSolaris since Oracle pulled the rug. Anything to foster Illumos for that matter, a lot of the distributions out there don't get nearly enough love. (OpenIndiana, Tribblix)

I liked it personally but it's cumbersome for everyday use. Sad BSD doesn't get the same love as Linux. Is there a way you can combine Debian with a BSD kernel instead of Linux?

>>71889 Literally: https://wiki.debian.org/Debian_GNU/kFreeBSD Spiritually: https://en.wikipedia.org/wiki/GhostBSD https://en.wikipedia.org/wiki/MidnightBSD Ten years ago: PCBSD Of general interest: https://www.dragonflybsd.org/

Apple raped BSD and thats one reason its not widespread because of Apple, AT&T and the fuarrrking penguin. fuarrrk you steve jobs you were a piece of soykaf.

I am using FreeBSD right now with Sway (Wayland) and Firefox. Most of everything works, seems like there's a million ports available, especially in the latest (unstable) repository. It makes an elegant and unstable desktop. I have not really used it for servers, I would just go with OpenBSD for that (which I have also used as a desktop, also went well. The drivers need improvement though).

>>70909 I haven't used any BSD, as all BSD's are proprietary software due to how hundreds of core BSD files have no clear license (and therefore are proprietary thanks to the Berne convention), plus every single one adds extra proprietary software under proprietary software licenses. When it comes to how proprietary, "Free"BSD is one of the most proprietary even.

What actual advantages are there to FreeBSD over popular Linux distros? >>72109 What's the actual difference between GPL and BSD licenses?

>>72186 >What actual advantages are there to FreeBSD over popular Linux distros? Smaller codebase and jails (https://docs.freebsd.org/en/books/handbook/jails/). Also, ZFS is much better than BTRFS. >What's the actual difference between GPL and BSD licenses? GPL is a copyleft license. It means that GPL license sticks with derivative forks and added new code, too. BSD license means that the added new code can be proprietary/non-free but you can't delete the BSD License from original files. You need to read https://www.gnu.org/licenses/copyleft.html and https://www.gnu.org/licenses/license-recommendations.html to understand this issue fully. >"Copyleft is a general method for making a program (or other work) free (in the sense of freedom, not “zero price”), and requiring all modified and extended versions of the program to be free as well."

>>70909 I use OpenBSD on my thinkpad, the experience was much better than other BSDs and pretty much on par with Linux, except for worse software compatibility. I'm guessing it's because the devs actually run OpenBSD on their laptops which are also mostly thinkpads. I really like DragonFly BSD and use it on my server, the filesystem (HAMMER and HAMMER2) is just so good and it has some other cool features like swapcache that can give a server a big boost. I really respect Matthew Dillon for his approach with DFly

>>72186 >What's the actual difference between GPL and BSD licenses? There are many different licenses in the GPL family and the BSD family, so that question requires quite a few paragraphs to answer. TL;DR: The BSD family is about gaining the maximal amount of popularity so the author(s) can get c(h)ucked the maximal amount and the GPL family is about giving freedom to all users of the software. All licenses in the BSD family are c(h)uck licenses, meaning proprietary software developers can c(h)uck over the author(s) and millions of people by taking it and turning it into proprietary malware (with the only requirement usually being trivial attribution, but even that isn't really required, no c(h)uck licensor has ever enforced the attribution part as far as I am aware). It appears that such licensors actually enjoy being c(h)ucked - one example is MINIX. If I remember correctly, Intel asked the author to help them make a special version that used the least possible amount of memory (without telling him what for), so he helped them and then they proceed to use it as the OS for their Management Engine frontdoor - leading to 500+ million people being c(h)ucked and he didn't even know about it for years or get paid. Further details (incorrect in some parts): https://lukesmith.xyz/articles/why-i-use-the-gpl-and-not-c(h)uck-licenses/ - remove the (h) to get the link, since language enhancer. People that license under the BSD family appear to be looking for maximum popularity at all costs, thus they effectively beg proprietary software developers to take it and incorporate it into their proprietary software, even though the fact of the matter is that is is never a good thing for proprietary malware to be better at taking the freedom of users. Some BSD-f licensors claim that they license pushover to give the maximal amount of freedom, but if you look at the end result (if the software is any good), you soon see that very few users of the software end up with freedom - most of them have their freedom better taken away with the software. Licensing BSD-f is about giving proprietary software developers the power to take the users freedom away and therefore maximize the amount of c(h)uckery for such developer and everyone else: https://www.gnu.org/philosophy/freedom-or-power.html The actual licenses themselves are also less free than the GPL family, as the BSD-f overstep copyright law and place conditions on usage unlike the GPLv3: 2-clause BSD: "Redistribution and ****use**** in source and binary forms, with or without modification, are permitted provided that ****the following conditions are met****" GPLv3: "9. Acceptance Not Required for Having Copies - You are not required to accept this License in order to receive or run a copy of the Program." The original 4-clause BSD also has a clause that prevents combining more than say 3 separate 4-clause pieces of software into a project: https://www.gnu.org/licenses/bsd.html Every BSD license I've read is also legally questionable, as every single one of them doesn't account for the many legal problems that plague software - which has mainly been done simply so the licenses could be very short as far as I can tell. Meanwhile, the GPL family (https://www.gnu.org/licenses/licenses.html) is designed to ensure that the 4 freedoms are respected (https://www.gnu.org/philosophy/free-sw.html#four-freedoms) to all users of the software. The way this is achieved is by using copyright law against itself to ensure freedom instead of taking it away - "copyleft". The goal is to allow anyone to run the software for any purpose, make changes to it and share modified and unmodified copies - even for commercial usages, but not allow anyone to take these freedoms away from any user - thus forbidding is forbidden. Due to the required legal wrangling to pull this off, the GPLv3 is much longer than the BSD 2-clause, but it has been been very carefully written with the assistance of many people and a lawyer and as a result the legal validity is rock solid. rms and the FSF originally released a bunch of per-project licenses for GNU software, but soon determined that everything should really be under a single license if possible and therefore made an amalgamation and published the GNU General Public License version 1. After a few years, it was evident that GPLv1 had many problems, so the GPLv2 was published with fixes - compatibility with the previous versions wasn't a problem, pretty much all authors decided to license under the -or-later option rather than -only - as a result, no software that's regularity used anymore remains under the GPLv1. After a number of years, it was determined that the GPLv2 had a few problems and parts that people didn't understand, so the GPLv3 was written by rms and his lawyer, except this time, a number of draft versions were published publicly so the community and also any businesses could point out problems with the license (which was pretty uncommon, as most licenses are only ever written privately). But of course, few people actually read the GPLv2 and GPLv3 or the faq (https://www.gnu.org/licenses/gpl-faq.html) and take incorrect online comments at face value, so most people don't understand what the GPL family requires and is about. At the same time, the Lesser GPL sub-family was released, which was the Library GPL for versions v1 and v2 and Lesser GPL for versions v2.1 and v3. The Library GPL name was originally chosen as it was only really intended for certain subroutine libraries - but it was too common for authors to read library and then use that license for their unparalleled library, so Lesser was chosen to note that it's a variant with lesser defenses for freedom. Most GNU libraries are under the GPLv3+, as they implement features that don't exist in any proprietary software (so only those part of the free world or willing to join it can convey it), but some are under the LGPLv2.1+, for example; glibc. While the ordinary GPL requires that all versions and derivative works remain free software, the Lesser GPL is a retreat of sorts, where 100% freedom is somehow counter intuitive. When glibc was being written, it was seen that proprietary libc's were a dime a dozen and it would really be better to allow anyone to use the libc, including the rare free, but GPL-incompatible licenses and even proprietary software, so a proprietary libc wouldn't cement itself onto GNU/Linux (as I guess they saw that many Linux developers couldn't help themselves but to do things that take the users freedom if convenient). The retreat is also measured - the library itself must remain free and the user must be able to make modifications to the library and install their own versions and reverse engineering of the non-LGPL'd part for the purposes of debugging modifications to the LGPL'd part must be permitted also. There's also the Affero GPL sub-family, which attempts to solve the problem where someone installs the software on their server and turns it into SaaSS: https://www.gnu.org/philosophy/who-does-that-server-really-serve.en.html The AGPLv3, on top of the GPLv3's requirements, requires that if you use the software to do another users computation and make a modification to the software, you need to share the source code to any users that access it over a computer network, so those users are able to host their own server and use that instead if they wish. This clause was going to be in the ordinary GPLv3, but SaaSS businesses complained, so it was decided to put that clause in the AGPLv3 and make the GPLv3 compatible with the AGPLv3 both ways. Thanks for reading my blog post.

>>72286 It seems BSD guys are more interested in recovering the market share they lost to Linux above all else

>>72291 >It seems BSD guys are more interested in recovering the market share they lost to Linux above all else I'm not sure what you're on about. BSD is a collection of proprietary OS's, while Linux is a proprietary kernel that is useless on its own (GNU Linux-libre is a free version of Linux meanwhile). All BSDs currently have more users than they have had in the past and I don't see any competition, as it's not like "Free"BSD users can't run GNU/Linux as well and vice versa - with the help of GNU autoconf, most software compiles and runs just fine on both even. "Free"BSD was initially released on 1 November 1993, which was a few months after GNU/Linux became increasingly convenient to use and therefore only the opposite of "losing market share" (whatever that means) could possibly happen. Sure the Berkeley Software Distribution predates GNU by a few years, since they were developed by universities from donated A&AT Unix source code, but GNU was the first free software OS, after over a decade of only proprietary OS's being available for recent computers, while no BSD has been free software yet.

>>72292 Linux is not really proprietary though. I don't like when people put blame on kernels instead of hardware manufacturers.

>>72186 >>72286 OK so the (2- and 3-clause) BSD licenses are < 0.3k words while the GPL2 or 3 is 2-3k words. It's probably faster to just read them all than to read these posts. The main difference between licenses BSD and GPL is that GPL requires derivative works to be also licensed under the GPL. BSD does not. Both state you must reproduce the copyright notice (includes the name of the holder) and there is no warranty. In the case of BSD that's about it. Otherwise, the GPL then goes on... it's relatively wordy and lengthy, and wants you to fill your stuff with GPL advertisements. If you want to prevent your competitor to use your code then GPL helps you. If you just want your code to be out there with a license that fits on a napkin, then BSD helps you.

>>72295 >Linux is not really proprietary though. I don't like when people put blame on kernels instead of hardware manufacturers. The Linux developers have explicitly decided to include proprietary software in tree, which makes Linux proprietary software. One example is: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/arch/powerpc/platforms/8xx/micropatch.c which contains software, as if I extract the contents of patch_2000[] and run cpu_rec, I find CPU instructions. On the top there's SPDX-License-Identifier: GPL-2.0 but that's not the license, as Motorola certainly didn't release such software under the GPLv2, as there's no corresponding source available. There's also plenty more, which is pretty bad, considering that all of such was meant to be moved to "linux-firmware" - but those Linux developers refuse to accept patches removing such proprietary software, even though such is clearly not used anymore, as nobody has moved it to "linux-firmware", so it really should have been removed by now on purely technical grounds alone. "linux-firmware" is arguably full of derivative works of GPLv2 software, but that's a different topic entirely. >>72305 >It's probably faster to just read them all than to read these posts. Yes, which is why I pointed out that you should probably just read them or the GPL faq. >Otherwise, the GPL then goes on... it's relatively wordy and lengthy, and wants you to fill your stuff with GPL advertisements. No section in any license in the GPL family requires advertising the GPL family. What's required in that the legal notices are left intact and if you modify the software, you indicate that the software has been modified, so if you modifications causes defects, those are attributed to you and not the original author. There is a recommendation that each nontrivial file has a suitable license header stating what the license is, so people can copy individual files without ending up with the files looking like they're proprietary, as the software and license is far too easy to separate without a suitable license header. Actually, the 4-clause BSD actually requires that you advertise its usage by including a statement as to which group authored the software in advertisements. >If you want to prevent your competitor to use your code then GPL helps you. Nothing in the GPL family prevents your competitor for using your code, for any purpose, your competitor is just prevented from proprietizing the software and thus gaining a proprietary benefit over you.

>>70909 My opinion of it is waning. I could not get WiFi to work properly, and suspend-to-ram is broken on all laptops I've tried (Thinkpads T61 and T500). OpenBSD is better, the developers eat their dogfood, so you can expect that things will just work on popular hardware.

>>72295 the official release of linux is proprietary, linux-libre which is maintained and released by the free software foundation is free software however

What's the best desktop environment for FreeBSD? I hate KDE

>>72597 Just use a WM. IceWM imitates Windows95 and it comes with a bar and a clock.

>>72597 Xfce works great, very functional and stable. As stated in >>72100 I use Sway without any desktop environment packages as those usually don't integrate well (or work at all). Why don't you like Plasma?

>>72612 Too much clutter and shiny buttons everywhere. I like the plainest, simplest, desktop environment. I have the attention span of a 5 year old so the simpler and cleaner something is the better. I love the simplicity of something like TailsOS. Maybe I should just stick with command line interface? I'm seriously contemplating doing everything in the terminal to cut down on distractions. It'd actually force me to learn how to use a computer too.

>>72613 >I'm seriously contemplating doing everything in the terminal to cut down on distractions. Do this. I really only have issues with mainstream desktop UNIX (Plasma, GNOME, ...). I may sound autistic, but something is always wrong, wrong enough that I can't use a normal desktop environment for more than a day. The million lines-of-code applications will crash and have broken features, not integrate well, etc. I find that using a window manager and terminals can be a lot more efficient than floating stuff, too. > It'd actually force me to learn how to use a computer too. I guess it would, but I'd say moving to a more linear operating system like FreeBSD will do that anyway, since you'll still have to maintain your system, no matter the desktop. Hopefully no soykaf graphical package managers and bullshit come to FreeBSD. > Too much clutter and shiny buttons everywhere. I don't think Plasma is cluttered by default, or often is, but you definitely can make it. I still understand what you mean when compared to a window manager. > I love the simplicity of something like TailsOS. I am guessing you are referring to GNOME. This is available, and semi-working on FreeBSD, but it will run terribly, as, I believe, Mutter is not properly ported to FreeBSD and has really bad graphical acceleration. I got it to run and it seemed to work well, minus the 15 frames per second I was getting.

>>72612 >Why don't you like Plasma? NTA. Aside from the frequent visual glitches I experienced, I think its too unopinionated. There's no vision to it. The devs even added a gnome style app launcher for people... who want gnome's interface... without using gnome? So you end up with something that can be finagled into a soykaftier version of windows, or a soykaftier version of mac. I'd rather have one good option than many bad ones.

>>70924 Anon runs a darknet site:

I hate KDE because its too customizable and I spend hours tweaking it trying to reach impossible perfection. It's so distracting. I need something extremely basic or my ADHD brain will just procrastinate and fixate on pointless soykaf.

>>72597 Lumina. It was made specifically for FreeBSD. Free from dependencies on redhat and freedesktop cancers.

>>71893 >Of general interest: https://www.dragonflybsd.org/ Thanks for the link. I never heard of DragonFlyBSD before. >>71893

>>70909 I have used FreeBSD to run Tor Relays, Mail Servers, and DNS. Jails are not as easy to use as Docker containers. ZFS is easier to use than BTRFS I dislike pf, compared to iptables/nftables, but people claim to love it. FreeBSD makes it incredibly easy to rebuild the OS from source in a way that Linux doesn't (including Gentoo). Documentation and technical know-how are generally lacking (despite FreeBSD Book), compared to Linux Materials. However, Linux routinely changes and removes control from you (IE. systemd take over of DNS resolution). I would consider using FreeBSD for a workstation but for Qubes is so good at automating VM creation.

>>73161 Personally, I wouldn't bother with DragonflyBSD, unless you know *BSD very well and you are proficient at C programming (including POSIX stuff). The community is just so small. Out of all *BSD operating systems, only FreeBSD and OpenBSD are good for a regular user (who knows basics of using terminal).

>>70909 I recently read article about bad defaults in FreeBSD base installation: https://vez.mrsk.me/freebsd-defaults.html >they used to re-add harmful/obsolete features back to OpenSSH >they have 3 different firewalls, including a fork of OpenBSD's pf. But they haven't synced their pf code with OpenBSD upstream since $forever. >for some reason, they use a worse NTP daemon by default, instead of just using OpenNTPD (it just works) etc. I used to be in favor of FreeBSD but it seems that OpenBSD is more sane by default (the problem is that they don't have a fast/modern file system or as good performance). >>73161 DragonFlyBSD used to have the best multi-threaded performance (IIRC, even better than Linux for a while). HAMMER2 is also very nice file system (but there is nothing wrong with ZFS or XFS or just ext4 either.).

It's just worse Linux (and I say that as an OpenBSD contributor and advocate). FreeBSD tries too hard to be Linux, but fails at it with it's corporate look and design. Compatibility with every big corpo out there, not even just ideologically, but also on a technical level. OpenBSD is not Linux and proudly so. >inb4 FreeBSD has Linux compatibility layer and wine But why not run Windows or Linux at this point if you need those programs. Imagine arguing for an OS, because it has some soykaf compatibility layer. >>70968 Licensing is something bad, no matter if it's copyleft GNU licensing or Microsoft EULA. OpenBSD's license is the best attempt at trying to distribute a free OS, without the regional restrictions imposed on the public domain in many countries.

>>74564 >But why not run Windows or Linux at this point if you need those programs. Are you seriously asking why not run Windows? It's overflowing with spyware and builtin bloat. A free as in freedom operating system that can run software compiled for whatever other os in a safe and stable manner is something I think us computer nerds can greatly benefit from. I'm not saying FreeBSD is that, far from it, but the concept of being compatible with software is great.

>>74594 Running non-free bloated spyware ontop of a non-free bloated spyware OS is not the colossal crime against nature you seem to think it is. >>74564 is correct, run windows applications on an isolated windows install (behind a well configured firewall). That will give you better compatibility, privacy and security than running it on some unsandboxed emulation layer alongside all your personal files. Especially FreeBSD which doesn't even make an attempt to be a secure OS.

>>74595 >Running non-free bloated spyware ontop of a non-free bloated spyware OS is not the colossal crime against nature you seem to think it is. An open source system that can run any program in a sandboxed environment can greatly reduce the ability of the spyware to spy. You seem to be suggesting an all-or-nothing approach to preventing being spied on

>>70909 FreeBSD will drop support for 32-bit hardware in the future OpenBSD, NetBSD and Void Linux (+ many other GNU/Linux distros) still have support 32-bit HW. I don't know how FreeBSD's decision will affect DragonFlyBSD, since dports is basically a copy of FreeBSD ports tree with DragonFlyBSD specific patches. But there has been talk about potentially switching to RavenPorts. >https://lists.freebsd.org/archives/freebsd-announce/2024-February/000117.html

>>76293 I don't know how FreeBSD's decision will affect DragonFlyBSD, Wasn't Dfly always AMD64 only?

>>70909 I tried to cpy ethernet frames from one interface to another in freebsd∆pfsense and it seems freebsd networking stack is not suitable for that .. I think this can be safely generalized for all of its components ...

>>76790 <cpy ethernet frames from one interface to another lainon, this is called bridge-ing, and both Linux and FreeBSD can trivially do it. On FreeBSD (man 4 bridge). ifconfig bridge0 create ifconfig bridge0 addm wlan0 ifconfig bridge0 addm tap0 Frames can now be copied from tap0 to wlan0. Linux has brctl and ipng utils. ip link add type bridge br0 ip link set up dev br0 ip link set dev ath0 master br0 ip link set dev tap0 master br0 Creating Virtual Bridges is table stakes for an Operating System now because of virtualization. Hypervisors like Virtualbox do this automatically to create virtual networks for you.

>>77560 in my case it was more about copying the frames literally , doing whatever post processing to them and injecting them into another interface. Bridging is layer 2 networking , I wanted full frame manipulation capabilities .. BPF is the way to go but that means writing your own code ...

>>70910 > Docker FreeBSD is where Jails came from. There would be no Docker without it. FreeBSD (and the BSDs in general) are where the ideas come from and are the most pure implementations of those ideas.

>>77644 > it was more about copying the frames literally , doing whatever post processing to them and injecting them into another interface. I might be stupid, but copying literally sounds not like doing post processing. I think this matters because with bridging, you are getting literal copying (provided the technologies are the same), the virtual bridge will copy to the virtual switch port that has device with correct MAC or flood. If you want arbitrary manipulation, nothing except code can give you *arbitrary* manipulation. It sounds to me like you want a TAP interface ( man 4 tap ), which has a master connected to a program that can preform arbitrary changes to frames. As you point out, if you are on Linux, you can do this kernel side with eBPF (XDP? ). FreeBSD should support it, as should every other OS, because TAP interface like functionality is essential for VPNs and for Virtualization. Qemu/OpenVPN both use for instance. Upshot being, its just not accurate to say that FreeBSD can't copy frames. Although you might say that FreeBSD doesn't have a programmable kernel the way Linux does with eBPF, and you prefer this solution.

>>70944 Also, Research UNIX refers to ATT Unix. BSD and SYS V both forked from the Seventh Edition of ATT Unix. Linux was based upon this , but shares no code. ATT continued to produce interesting Operating Systems including Plan9 and Inferno after.

Is there a difference between ZFS and OpenZFS or is it just different names for same thing?

>>78556 ZFS is name of a filesystem. OpenZFS is one of the implementations of it. I think Oracle has its own ZFS implementation too.

>>70916 If you are just now quitting windows you shouldn't jump to FreeBSD, i think you should start out with Linux Mint and then once you become comftorable with that you can start distrohopping.

>>78623 There is GhostBSD though which polishes FreeBSD to be more easily used for beginners.

>>72292 >while no BSD has been free software yet. >BSD is a collection of proprietary OS's Why do you say that?

>>77790 something on the level of TC (linux) is missing from *BSDs ...

insecure because the security team is incompetent

>>79427 Why do you think they are incompetent and where is the security concerns?

>>79427 >insecure because the security team is incompetent Seems a little provocative, and you should probably elaborate. nb4 OpenBSD. OpenBSD doesn't have a MAC framework, doesn't believe in virtualization, and is the only OS still trying 'secure by correct design' as a practice.

>>79471 that's the point, almost nobody out there is competent and those who are cba to waste their time implementing "correct" software. software is meant to work and be convenient to use, then comes security. even in military. utility first, reducing losses next. so bsd is a futile effort. even on servers people have been reporting that linux performs better than tuned bsd. security doesn't exist. privacy either. you're just being autistic

>>79472 >Software should look to the military for function >Security doesn't exit >Your being autistic Insane take. Security does exist, its all about risk posturing (acceptance, avoidance, mitigation, transference) and decision making under uncertainty. Whats constitutes good security thinking in software has nothing to do with the military, and I am not even sure your thinking about the military is correct. This matters because there are billions of dollars of economic value at stake, proven out via ransome-ware, and perhaps the entirety of our society in a speculative "cyber Pearl-Habour" scenario. Obviously there are no easy answers, but security doesn't exist just use Linux bro, its military is actually autistic.

>>79479 you have said nothing in an entire post so why did you bother at all? security is a measure to prevent your enemy from winning you, not from damaging you. preventing damage doesn't make sense if there's nothing of value behind your security. bsds are reportedly (i didn't confirm to be honest) gradually looses its positions to linux in enterprise and it's barely usable on desktop. now what's the value of all their security when at this rate nobody is even going to be really interested in breaching it in a decade? security doesn't exist in itself, it is a weapon to be used in the heat of battle if you're afraid of loosing your data you should probably chop off the wire for good

I use FreeBSD as my main computer, albeit headless: it hosts a samba share, runs rtorrent in a tmux session. main thing it does is centralize my computing into 1 box, i generally access it via emacs tramp (ssh) or plain old ssh. out of the box the coreutils are a little more privative with bsd variants of sh, grep, etc versus gnu's on debian. the file system is more organized, and there's a ports system for ease of compilation. additionally the service manager/init is pretty much shell scripts. most system configuration for typical sysadmin tasks are done through 3 files in /etc/. zfs is also seamlessly installed for the OS drive, overall not a bad unix with the caveat that you're dealing with a unix/unix like.

>>79479 >I am not even sure your thinking about the military is correct. In theory the military pays the ultimate price for getting things wrong. In reality that only applies to a small subset on the front line and the rest is just as corrupt and incompetent as any other government organization. >>79480 >security is a measure to prevent your enemy from winning you My enemy is not your enemy. And the military's enemy is definitely not your enemy. That's what the post you're responding to means by risk posturing. >bsds are reportedly (i didn't confirm to be honest) gradually looses its positions to linux in enterprise BSD lost that fight a long time ago. In the early 90s there was a massive copyright lawsuit between AT&T and the original BSD project which took years to resolve. By the time it was over everybody had already started to use linux and the new open source BSD forks have been trying to catch up ever since. For a long time the only thing FreeBSD had over linux was zfs being lightyears ahead of btrfs but they don't even have that anymore. >>79471 >Seems a little provocative, and you should probably elaborate Not that lain but BSD security is a joke telnet buffer overflow (2023) https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=270263 ping buffer overflow (2022) https://www.scmagazine.com/brief/freebsd-addresses-critical-ping-vulnerability They didn't even bother to add aslr until 2019. Even fuarrrking Windows had aslr 20 years ago. https://svnweb.freebsd.org/base?view=revision&revision=343964 NetBSD security audit finds 'hundreds' of bugs (2018) https://www.theregister.com/2018/05/29/netbsd_code_security_audit_finds_hundreds_of_flaws/ >inb4 OpenBSD. OpenBSD doesn't have a MAC framework, doesn't believe in virtualization, and is the only OS still trying 'secure by correct design' as a practice. OpenBSD tries hard but they're not as invulnerable as fanboys delude themselves into believing. Especially since as you said they have no sandboxing technology, you just have to blindly trust that all code you run is safe and benevolent. openbsd heap underflow (2024) https://www.supernetworks.org/CVE-2024-43688/openbsd-cron-heap-underflow.txt openbsd heap overflow (2022) https://blog.quarkslab.com/heap-overflow-in-openbsds-slaacd-via-router-advertisement.html

>>79577 >openbsd has no sandboxing They literally invented chroot >No virtualization https://www.openbsd.org/faq/faq16.html

>>79579 >They literally invented chroot https://en.wikipedia.org/wiki/Chroot >The chroot system call was introduced during development of Version 7 Unix in 1979. https://en.wikipedia.org/wiki/Openbsd >Theo de Raadt created OpenBSD in 1995 chroot is not a sandbox, root inside a chroot is root everywhere https://infocondb.org/con/deepsec/deepsec-2015-deepsec-no-9/chw00t-how-to-break-out-from-various-chroot-solutions >No virtualization https://marc.info/?l=openbsd-misc&m=119318909016582 >Supported guest operating systems are currently limited to OpenBSD and Linux. As there is no VGA support yet, the guest OS must support serial console. Wow, maybe in another 20 years it will be able to do something useful.

>>72059 Apple and Netflix are the only reasons BSD is still relevant

>>80505 BSD stuff is the base of various gaming consoles OS are built on and likely some digital media players among other things.