My home server has been running FreeBSD for ten years now, and it has never let me down. Except for one time I got fresh with /dev/speaker and triggered a spontaneous reboot (I don't know if it's FreeBSD's fault or the hardware, though).
I delayed upgrading to 15.0 after it was released, but last weekend I finally did it, and it left me wondering why I hadn't done it sooner, because it went quickly and smoothly.
Is there anything FreeBSD can do that, say, Debian cannot? Probably not (at least I cannot think of anything). When I set up the server, ZFS was a huge selling point, but I heard that it works quite well on Linux, these days. But I appreciate the reliability, the good documentation, the community (when I need help).
There are various niche applications where Debian or any Linux are worse than FreeBSD.
For example the support for magnetic tapes and for a few other SCSI peripherals is better in FreeBSD. The Linux utility for controlling a LTO tape drive lacks some important options that the corresponding FreeBSD utility has.
I have a tape drive, and to be able to use it like I want I had to move it to a FreeBSD server.
Some years ago I was using a surveillance camera that was much easier to use in FreeBSD than in Linux, if you wanted to record good quality video and audio. I have not tried more recently to use such cameras in Linux, to see if now the recording quality is better.
So while there are more hardware devices that have better support in Linux than in FreeBSD, there are also devices with better support in FreeBSD than in Linux.
However the main reason why I use FreeBSD on many of my servers is that I need much less time for their administration than for Linux servers. In my experience, Linux servers need much less time for administration than Windows servers, and FreeBSD compares to Linux like Linux to Windows.
I have FreeBSD servers that I have not touched for years, and they have worked 24/7 with no downtime and no rebooting, and this includes servers connected directly to the Internet, which implement firewalls, routers and various services, like NTP, DNS servers and proxies, e-mail servers, web servers and proxies etc.
Cameras? I suppose the world still has some weird cameras that need proprietary/weird drivers, but for all intents and purposes: USB cameras are UVC and work with a generic driver, and IP cameras are OnVIF and work with ffmpeg. I can't imagine the latter having any OS dependencies as far as Linux/BSD/Mac/Windows is concerned. Quality is fine - I have a bunch recording 24/7 with high quality audio and video.
> I have FreeBSD servers that I have not touched for years, and they have worked 24/7 with no downtime and no rebooting, and this includes servers connected directly to the Internet, which implement firewalls, routers and various services, like NTP, DNS servers and proxies, e-mail servers, web servers and proxies etc.
Same. We've got qmail config files with 2006 as the mtime
> Some years ago I was using a surveillance camera that was much easier to use in FreeBSD than in Linux, if you wanted to record good quality video and audio. I have not tried more recently to use such cameras in Linux, to see if now the recording quality is better.
This example seems very hand-wavy. What camera?
A Logitech FullHD camera on USB, but I doubt that the problem was camera-specific. I believe that I would have seen the same behavior on any high-resolution USB camera.
In FreeBSD, the command required for recording was very simple and it worked flawlessly. In Linux, it was more complex and there were various stuttering problems at maximum resolution. I am still using those cameras, but I have not tried them again in Linux. In Linux they worked worse than in FreeBSD around 5 years ago, perhaps nowadays there is no longer any problem in Linux.
This was intended to be an example that you cannot know a priori whether a given device will work better on FreeBSD or on Linux. In general, there is a greater probability for Linux to have good support than for FreeBSD, but there are also counterexamples, so you cannot be certain which is better until you try both.
I am sorry, I have a hard time accepting this level of detail, acknowledging it was half a decade ago.
In a nutshell, you content that FreeBSD running on the same hardware as "a linux" performed better with camera operations. However, you did not specify even a specific camera model, or the interface(s) used to interact with the camera.
I have zero issue accepting that a BSD is better than a linux at things, pretending otherwise is foolish. However, this specific example isn't tracking.
> Is there anything FreeBSD can do that, say, Debian cannot?
If you asked the opposite (what can Debian do that FreeBSD cannot) I would have more to say and it would mostly be preceded by "I know FreeBSD is not Linux but ...". Whenever I need to do any sort of maintenance or inspection I have to look up the equivalent commands for things like `lsblk` and something nested in `/usr/etc/...` when I'm used to finding it in `/etc/` over every other system.
This is a consequence of both FreeBSD's reliability in needing very infrequent attention and my limited use-cases to use it. As a NAS it is great but I can't touch it without full-text search of all my notes on the side! Either way, no regrets about learning and relying on it after ~18 months so far.
Lack of good NFS support? When we benchmarked it last it was 10x+ slower than running on linux (ubuntu).
Also lack of collective mindshare. I use FreeBSD at work every since day and while I don't hate it, I do wish we just used Linux. There are more guides, tools, etc for Linux than for FreeBSD. Yes, as a comment in this sub-thread stated, jails exist but everyone knows docker, not jails. So even with jails apparently being better than containers, it doesn't really matter, there isn't the ecosystem there.
FreeBSD might be as good as this blog author makes it out to be, and maybe I'm "holding it wrong" (always a strong possibility) but I can't help but feel it causes more friction than I'd like, it's just "slightly" harder to do anything. In the age of LLMs I have to tell it (or put it in my system prompt) "I'm using FreeBSD" or it will be give me Linux advice. It just feels like death by a thousand papercuts.
>>maybe I'm "holding it wrong" (always a strong possibility)
Yes. You are holding it wrong. And it's obvious from your comment.
Lack of docker support? Docker is available on macOS through emulation yes but bhyve is a thing… so why not? :-)
That's why I don't use Linux. It lacks Jail support.
Docker is a concept resembling FreeBSD's jails that were introduced in year 2000, having much better isolation, much better security than Docker has had for a long time (perhaps even now jails are still superior to Docker).
Better isolation, better security, but far fewer gists and shared config-files shared ok the Internet for common tasks. Docker comprehensively wkn thr popularity contest, and is often the more convenient solution because of it, in a worse is better way.
ZFS on FreeBSD is first class. I had an old FreeNAS raid z5 array on 5x 500GB disks that I wanted to check 4 years after decommissioning the system. I put together a temporary machine with all the disks plugged in and without doing anything the live FreeBSD image found and configured the array. I was instantly able to look through the file system and even dump it to my current FreeBSD server with almost 0 effort. I was sold after that. These days I prefer to run small systems and basic services. I don't want webguis or docker images anymore.
Right. On my development workstation I use Arch and I'm always worried a kernel upgrade is going to break the ZFS module. For those that aren't familiar, ZFS isn't part of mainline Linux because of licensing incompatibility (and general distrust of Oracle).
On FreeBSD I know its always going to work.
>Is there anything FreeBSD can do that, say, Debian cannot?
ZFS boot environments.
One could install Debian's root on ZFS by following the OpenZFS documentation guide, combine it with ZFSBootMenu (or similar), but there won't be any upstream support from the Debian project itself.
The Nitrux Linux distribution is based on Debian and provides an immutable feature similar to boot environments, but you can't treat your immutable boot images the same way you can treat your mutable data like how you can with ZFS datasets on FreeBSD.
you can use snapper + btrfs and the end result is like `bectl`. However it not as simple/integrated as ZFS Boot environments on FreeBSD
On openSUSE Tumbleweed, it is. Each Upgrade creates two snapshots, one before, one after, and if anything goes wrong, I can boot into a snapshot where the world was still in order.
I have a higher opinion of ZFS than I do of btrfs, but FWIW snapper+btrfs has worked well for me on openSUSE Tumbleweed for ten years now, too.
The btrfs code quality seems less than ZFS, based on the reports I have read.
Last I heard (~8 years ago), the RAID-like functionality in btrfs was very unstable and crash-prone. The impression I got was that there was not a lot of interest in fixing this. Then bcachefs came and ... appears to have gone nowhere AFAICT.
The non-RAID part of btrfs appears to be stable. It's the default filesystem on openSUSE and SLES. But I don't think it's ever going to reach feature parity with ZFS.
btrfs is suffering from a lot of old bad publicity and some poor design decisions around RAID.
But by now it is a great file system if you don't go near RAID5/6. btrfs has its flaws (ZFS has its own flaws!). However:
- It's used a lot, especially by facebook and Redhat (on fedora)
- Gets a lot of testing
- Sees a lot of bug fixes
- Has a lot of features
I haven't read btrfs code but given that it is a popular file system and Linux code quality tends to be good in popular subsystems I would hesitate to say its code quality is worse than ZFS in any way.
btrfs is pathetic when it comes to performance. So no, thanks.
https://www.phoronix.com/review/linux-70-filesystems
In real world scenarios, where file based backups fail, one needs to add at least lvm.
And only than those benchmarks would be more interesting to me.
> Is there anything FreeBSD can do that, say, Debian cannot? Probably not (at least I cannot think of anything).
Stability of user interface and documentation.
What documentation does a distro even publish? I only ask because freebsd has the most solid documentation I've used of any OS I've ever encountered. I seriously doubt Debian has documented the linux kernel that well (which, tbf, would be an insane project to even attempt)
Debian has an installation guide[1]. I'd imagine all the major distributions do.
But it probably has to change a lot for every major release, because so many things change. FreeBSD major releases have changes too, but a lot of the user interfaces are very stable and so the documentation can be too. Stable documentation allows time for it to be edited and revised to become better documentation, as well as developing quality translations.
[1] https://www.debian.org/releases/stable/amd64/
I've been surprised at times by what's missing. There are are some strange omissions, from most recent memory around the Bourne shell builtin commands and make(1). I've had to go hunting in other BSD distribution manuals at times to find what I need. I'd say the GNU core utilities sometimes have better docs for their equivalent commands.
That said, for non-core utilities on Linux it's pretty hit-or-miss. The BSDs are generally pretty consistent in what they do offer, and that's what I love about them. Of course it's a different development model and it shows.
> Is there anything FreeBSD can do that, say, Debian cannot?
Yes. Emulate traffic latency using IPFW and dummynet[^1]. There is no Linux (or OpenBSD, NetBSD) counterpart.
The ZFS implementation is less buggy.
[^1]: https://man.freebsd.org/cgi/man.cgi?dummynet
That is not really accurate? Linux traffic control (tc, [0]) exists since Kernel 2.2. It can introduce traffic latency and a few other network conditions, like packet loss.
[0]: https://www.man7.org/linux/man-pages/man8/tc.8.html
Hmm kind of... I was referring to the fact that dummynet models pipes with a fixed bandwidth and centralized scheduler. Packets are released according to very high precision transmission timing. This means that serialization delay, queue buildup, and link behavior are simulated in a way that resembles real network conditions. Dummynet can provide a highly deterministic timing and queue behavior, which made it popular in networking research and WAN emulation experiments. TC cannot do that with the same accuracy.
I think much like other tools, think SELinux vs OpenBSD (unveil, etc) TC is more flexible (does more things) but there are _some things_ that can't do, and even for things both can do *BSD solutions are much simpler.
> The ZFS implementation is less buggy.
FreeBSD and Linux have been using the same implementation of ZFS for years.
You can emulate latency, packet errors, etc using netem tc [0] on Linux.
[0]: https://man.archlinux.org/man/tc-netem.8.en
What I really want is the Windows tool for that. Can't call it equivalent, because clunsy is way superior.
https://jagt.github.io/clumsy/index.html
My current home server passed 10 years in the autnum, but I've been running FreeBSD on servers since around 2000.
The main gripe is probably Docker and/or software depending on Linux-isms that can't be run natively without resorting to bhyve or smth alike that.
You could just use podman.
Theoretically yes, however still limited by how well the FreeBSD Linux layer handles syscalls. A year or so back I tried running .NET (just binaries, not via a container) since the port wasn't as far along as today and it crashed due to what I suspect was slight differences in signal handling defaults.
And this is part of the situation that's going to get worse, io_uring will become more popular in language runtimes and iirc it's not trivial to emulate via existing FreeBSD mechanisms (kqueue).
Iirc Mac docker uses xhyve (bhyve port/inspired) to run containers via Linux emulation, MS went for pv-Linux for WSL2, while FreeBSD has been "good enough" so far.
But I think that for containers it's either time to shape up Linux emulation well (It's ironic that WSL1 ironed out their worst quirks just as WSL2 was introduced, although that was without io_uring) or just add an option for Podman to have a minimal pv-Linux kernel via bhyve to get better compatibility.
Indeed, ideally we could get docker on FreeBSD using the same approach as is used on macOS — automatically run (one or more) Linux VMs under bhyve.
I wonder if FreeBSD ought to consider a WSL2-style approach to Linux binary compatibility, too.
Keeping the Linux syscall compatibility layer up-to-date has always been a resource problem, especially when syscalls depend on large, complex Linux kernel subsystems that just don’t map cleanly to FreeBSD kernel facilities.
Exactly the reason why I switched from FreeBSD to Debian, 25 years ago
[flagged]
Docker's what lets me spend more time using the software on my server than fiddling with it. I got the fiddling out of my system years ago, I just want shit to work now.
I don't really care about it per se but having a cross-distro unified daemon config & supervisor, package manager, and ability to cram every single important file into a file tree (again, using the same interface to achieve this with every daemon) that has only those files in it (making backups and restores trivial) and then easily verify that I got all the important files (destroy image -> re-create, does it still look good? Then I got everything) makes everything so easy. I no longer put off trying a new service until the weekend because it'll take an unknown amount of time that could end up being hours. Odds are I can have anything available in the official Docker registry (which is approximately everything, these days) up in five minutes flat to try it out, and may not even need any further modifications for it to be ready for (personal) "production".
I use Debian but don't even care, I haven't had to touch systemd once (thank god) and the only Debian-parts I even use are its ZFS, SSH, and Docker, with my ten or so actual user-facing services all just pulled and managed via Docker, ready to transfer to any other distro seamlessly, should I ever care to. Even Samba is under Docker (oh my god it is so much easier to configure for common use-cases this way).
(I would definitely be using FreeBSD on my server if I cared about anything other than Docker, though—I haven't actually liked Linux for about fifteen years now)
I've fixed too many linux-isms in code manually over the years, pkg/ports does what it should but since *nix tradition relies on hardcoded paths I often wanted out-of-tree builds for various software.
And that's the thing, as I grow older I feel more and more that I just want the parts of computing that I don't want to _care_ about to be stupid simple.
If I'm doing a program needing a recent version of some language that doesn't have a FreeBSD port yet and some database behind it, I don't really want to configure it all manually because I don't particularly care for porting the runtime or managing the database (that isn't exposed to the outside world anyhow).
This is stuff where I don't want a large CI pipe or other management (or needing to remember to upgrade the packages if I upgrade the hosting OS).
Stuff like this is why "the clouds are winning", friction should be linear depending on the effort I want to put into managing something.
But going for real HW or even VPS a places an "upgrade tax" on me because I can't just let non-public services like an isolated DB just ride-along over major versions (maybe Jails with isolated userlands could be an option, but that becomes painful instead when needing newer versions of the application behind the veil).
docker and flatpak/snap are _extremely_ different tools with very different purposes.
> I delayed upgrading to 15.0 after it was released, but last weekend I finally did it, and it left me wondering why I hadn't done it sooner, because it went quickly and smoothly.
I haven't done that yet because I think I'd want to switch to pkgbase but that makes me nervous. Did you go with that option or continued to use the sets?
Why not create a boot environment and try out FreeBSD 15 with pkgbase ? If the experiment is successful just make the boot environment the default otherwise throw it away...
With such powerful tools I find it fascinating that FreeBSD users are not more willing to experiment !
To be honest I've never tried boot environments. I know they are a thing, and I have my whole setup on ZFS, so maybe that's the perfect use case.
I haven't switched to pkgbase. Yet. I don't intend to for the time being. I set up a VM to test it, but I haven't gotten around to actually testing it.
FreeBsd is Systemd free.