Any FreeBSD Users here?

I just listened to @bryanlunduke episode about FreeBSD.

Are there any FreeBSD users here?


Not this guy. :slight_smile:

Not currently, but I’ve done a little bit with BSD (ensuring software written for Linux also runs on BSD)… Also used BSD servers.

Whenever I look into BSD, it seems mainly for sys-admins: file-storage and networking are awesome… other stuff is fine, but not really better then Linux from a user-perspective. This interview re-enforced that.

As a user & developer, a lot boils down to marker share. If you have some random problem - what are the chances someone else had the same problem? Can you find a fix in 5 minutes searching online?
For Linux this has improved a lot. Using BSD as a desktop/workstation feels too much like using a niche within a niche.

To the last time I used FreeBSD, I tried to build some QT5 application I’d ported to Linux and was curious if it would take much extra effort to run on FreeBSD - it failed with a strange error. From what I could tell it turned out to be some interaction with how CMake inspects the system’s QT5… spent in hour or so trying to fix (it wasn’t some trivial missing file or wrong path error) I gave up. I could have reported it, I just wasn’t that invested.

As for Bryan’s interview, thought he did a good job asking some slightly uncomfortable questions, reasonable given Linux’s dominance. The case for open code in IOT was well made too, While George had a point about binary-blobs, I found it a bit of a cop-out - something to the effect of “if there are some binary blobs in the system, everything might as well be closed-source”.

It seems the effect of a permissive license has played out, that they are used by others who arent investing so much back in the FreeBSD eco-system, a while back some of the lead FreeBSD developers were hired by Apple and from what I can tell just stopped contributing to FreeBSD. While they can be proud they’re used by Apple and Sony, these companies are not big supporters of FreeBSD from what I can tell. The ZeroMQ author wrote an interesting article on this topic.
On the other-hand, iX and NetFlix do contribute back, nevertheless IMHO its significant that some of the biggest users havn’t. 2017 recently posted their videos online, Benno Rice (a FreeBSD developer) gave an interesting talk on some of the problems with FreeBSD, found it quite insightful, while FreeBSD developers are proud of their stability and development model (where everythings maintained centrally), It’s nice to see some honest shelf criticism too.


That linked article is really interesting in the context of the recent licensing discussions.

I also find that linked article from @ideasman42’s post very interesing…but to be honest, I must admit that I just don’t really understand it :frowning:

I mean, about this Patrick Guy…why was it stupid to choose the BSD license?
How would things have been different if Patrick had chosen GPL over BSD?
Or was the stupid thing that he didn’t react by also relicensing his project to GPL?
Was he therefore somehow unable to benefit from the other group’s work?

I feel that it’s hiding from me in plain sight…
Could someone please explain?

OpenOffice is under a permissive license and was forked and re-licensed copy-left (LibreOffice), meaning LibreOffice can take changes from OpenOffice, but not the other way round.

Now imagine (for arguments sake) the OpenOffice developers were angry and complained that someone else “took their code”?
A reasonable response could be: “if you don’t want this to happen - don’t use a permissive license”.

This is just an example, I have no knowledge of how these projects interact.

1 Like

Ah, so they (or Patrick) choose not to switch to GPL
…in fact, they (he) could have done it, but chose not to.

so from the article:

[…]but then he soon realized that new patches coming from that project couldn’t even be merged back into his work!

I.e. it wasn’t technical problems but it would have forced him to re-license his work to GPL as well, which he did not want to.

OK, thx a lot :smiley:

I am actually using TrueOS, formerly known as PC-BSD, which is a “spork” of FreeBSD. I think the best description of the relationship between TrueOS and FreeBSD is it is similar to the relationship between Linux Mint and Ubuntu. I started dabbling with BSD as systemd started taking over more and more of the system, but it was ZFS and jails that that kept me interested.

I don’t know that I agree wtih this. One of the features of TrueOS is the concept of “Boot Environments.” The nice thing about a boot environment (“BE”) is that when you upgrade the system, it creates a ZFS snapshot and does the upgrade in that. Two of the practical effects of this is that upgrades don’t affect the running system. So you aren’t going to have an app start misbehaving because an underlying library gets upgraded. So it requires a reboot to get any changes.

The other advantage is that if you do reboot, and something goes sideways, you can completely roll back to an earlier snapshot, troubleshoot the issue, and upgrade again.

Quick story. I had a test box that was using the nvidia legacy video driver. One upgrade installed the nvidia drivers, which meant that after reboot, I had no X. I was able to roll back to the previous BE and fix the issue, then upgrade again.

Just my experiences. FreeBSD seems to be about where Linux was 8 or 10 years ago. I have found very few apps that I was running on Debian that I haven’t been able to run on TrueOS.

Full disclosure, I am a sysadmin.

Good point (I was considering ZFS more system-level than user-level, but your right about boot-environments).

While boot-environments are great, to gain a feature I almost never need, I’d then have to deal with many small annoyance that need to be resolved.

It’s not really BSD’s fault, but when you run into glitches simply by using a more obscure system (obscure as a workstation, its common for servers), these are paper-cuts that impact you on a regular basis. The chances you can build something thats just been released is much lower for example. Even if it eventually gets support.

Note that I tried TrueOS twice… first time the package manager stopped updating packages (it couldn’t validate a security key and was determined not to let me bypass security - no info online for how to fix)… second time I was under the impression I was installing a rolling release, but after some weeks didn’t see any updates (probably my fault… maybe it needed to be turned on somewhere?).
There were various issues trying to build software and having to use ports to get recent compilers is a bit tedious when you can access them as packages elsewhere. Also looked into PacBSD but it seems very under-developed.

Added to this, interviews I’ve heard with FreeBSD developers - often say they use OSX for laptop/desktop, FreeBSD on the servers… which doesn’t motivate me to be a trailblazer something the developers themselves aren’t interested in.

Mac user. :slight_smile:

1 Like

There have been more than one time when an apt-get update && apt-get dist upgrade on Debian/sid beamed me directly to the Planet of Unintended Consequences, and I wish I could wave a magic wand and get my system back. As it was, I had to go through and manually fix and reinstall packages.

I guess I come at it from quite a different perspective. I have been doing Linux for nearly 22 years, so when I started with Linux, it was far smaller than BSD is now. So I’m used to having to find workarounds. That said, I am sitting on my TrueOS box running KDE, with most all of the stuff I used on Debian/sid before. I did have to put my wife’s machine back on Debian because the TV app that she liked didn’t work on Free/PC-BSD, because of the differences between V4L and the way BSD does it, and she didn’t like any of the BSD TV apps.

That said, I have set up a number of appliances. I am running a FreeNAS box, and a pfSense box, and they are running like tops. I even had a drive fail in my FreeNAS, and it was as easy as it said on the tin. Get the offline the failed disk, replace the physical drive, then click “Replace Disk” and let it resilver. There was also no functionality loss from my IPtables box to the pfSense box. Everything including VPN works for a treat.

Remember that TrueOS was still in alpha or maybe early beta then, and is considered beta now. Especially back then, updates were…spotty. This is kind of a strange time for the project. They rebranded from PC-BSD to TrueOS, and went from being based on -RELEASE to -CURRENT (rolling release), So there is still a lot of churn.

That said, I have still had more issues with Debian and some of their critical and grave bug reports, which seem to have gotten much worse over the past few years…And I have been running Debian since 1999.

Well, first of all, there are quite a few OSX boxes at Linux conferences too…Now that said, I know that both Kris Moore (TrueOS’ lead developer and Allan Jude (ZFS/FreeBSD dev) run (or multiboot) TrueOS and/or FreeBSD.

The bottom line is that Mac (and to an extent windows) have infiltrated the open source community. I personally prefer the open source desktop; I used to run Enlightenment back in thhe day, and have been on KDE for years. And I wouldn’t say that the developers “aren’t interested” in the desktop. You have to admit that more and more of the desktops are making it difficult for the BSDs. For instance, Gnome has a given their undying love to systemd, and made it a hard dependency. None of the BSDs use systemd, which means they spend untold hours making it work for BSD by ripping out that dependency. That said, jave you taken a look at Lumina? Written by Ken Moore (Kris Moore’s brother), this is the first BSD-native desktop. I would have to say that it seems that there is at least some interest in the desktop.


I tested TrueOS/PC-BSD was less then a year ago, and feel a bit sheepish bringing this up problems with it, since should have really reported bugs and helped out… I just wanted to get a sense of where they were at.

I’ve never used Debian, currently running Arch and somehow managed to avoid running into bad distro problems (except for one time!). TBH, not even sure I would use ZFS… It’s nice to have fast SSD access, and anything important gets backed up or goes into version control.

As for desktop stuff: I probably should have explained what I meant by desktop/laptop use - The actual desktop environments I’m not really interested in (I’d run the same tiling window manager on FreeBSD too), I was referring to the technology that supports the desktop as a focus for general development - Pushing forward with Wayland, general graphics card driver support… I realize FreeBSD can do most things. It just tends to be a bit behind.

While I realize TrueOS/FreeBSD are good, I would probably end up setting nearly the same user-environment I have on Linux - so it just seems too much effort to switch to a smaller niche with slightly worse software support. OTOH, a nice rolling release would be tempting to try, but more as a curiosity.

This is what I try to do, but sometimes I feel like the issues I bring up are dumb…Don’t sweat it. This is another thing I have found, is that it seems that the BSD community is more helpful and friendly than the Linux community. I was on the Debian User mailing list for over 10 years. Finally, in 2014, I had to leave because of the constant flame wars becoming more prevalent than actual helpful posts. The Arch community has a reputation as being elitist as well.

Actually, I think FreeBSD has an active Wayland effort, if Phoronix is to be believed…And I heard something similar on BSD Now the other week.

That is something that they are currently working hard on in TrueOS. They have pretty much completed the transition to OpenRC, and now are working on the video stuff.

That is exactly what I did with my TrueOS installation. I’m running KDE on Xorg with all of the same apps I had on Linux. But the reason I started dabbling with it is that the nonsensical changes to the way basic tools work, and the systemd team’s stated desire to “own everything between the kernel and the application layer.” And then, the advantages of ZFS was what really kept me there.

That’s a shame, though I have the impression when communities get too big and their purpose is too broad - they stop functioning properly as communities, just as windows or OSX users don’t necessarily have much in common.

As for Arch community being elitist, while I didn’t have a lot of contract, I always found them to be reasonable (mainly dealt with their package maintainers though IRC and issue tracker).

Right, I’m aware of this, the point is that its behind and that this isn’t a new thing. Often some technology comes out on Linux, and its some time before the BSD’s get it… and even when its supported. The support may be worse simply because not as many people are using it and reporting bugs.

The issue I pointed out before is an example of this: FreeBSD supports QT5, but I couldn’t get a QT5 application past the library configuration stage because of some non-trivial problem with CMake.
(If someone reading this post is curious, this is the QT5 application I was trying to build).

Which means he’s more of a FreeBSD user than someone running Linux or Windows :wink:

I have heard on more than one occasion, that the response to someone’s question would be RTFM! That is what I meant by elitist.

It depends. There are some packages (beyond ZFS) which there are versions at least as new of not newer in BSD. Lumina comes to mind, but that is a DE that was developed in BSD that has been ported to Linux, rather than the other way 'round.

Now having said that, it is a growing problem that Linux applications are becoming more and more Linux specific. For instance, I heard that the Budgee desktop is abandoning GTK/Gnome for QT for their desktop because of the random changes that the gnome team makes.

It’s even worse for the BSD community, as more and more apps make systemd, or cgroups or whatever Linux-specific knobs hard dependencies. This means that BSD devs have to spend untold numbers of developer hours either going through and ripping all that stuff out of there, they have to fork the entire project., or come up with a BSD-licensed equivalent. In any case, fix or reinvent the wheel…

Are there really many/any applications that depend on systemd?

Besides desktop’s and their bundled utilities.
Are web-browsers, image-editors, video-editors, office-suites… etc - that can’t be used without systemd?

Interestingly Jordan Hubbard (a FreeBSD/OSX dev) makes the case for having a systemd/launchd/SMF equivalent init system in FreeBSD.

There was some work to modernize FreeBSD under the project name NextBSD, but it seems to have stalled.

I’m glad you made that comment. I had always assumed that Linux was derived from Unix. But you made me look into the matter and found I was in error. That leads me to another question. Have you or anyone delved into Apple’s open source Darwin project? Or do you know of anyone that has?

Also curious, there seems to be surprisingly little interest in doing so, probably because it doesn’t include the graphics stack.

Q&A on the topic, looks like its possible at least.

There is also this project going on. Seems like early days for it. I was looking at Apple’s darwinos users page, but it doesn’t seem to be much. I tried the FAQ link to no avail.

Please respect our code of conduct which is simple: don't be a dick.