I run OpenBSD on a PC Engines apu2 as my main router. I have a 1 Gbit fiber going into a TP-Link converter and then directly into the OpenBSD router which completely bypasses the POS router provided by my ISP.

When I first started poking around with BSD I searched for tutorials online for the things I wanted to do. As I often do when I want to learn something. Early on in this process I came across a suggestion to just read the man pages (RTFM!). So I stopped searching and just read the official documentation. I was surprised how it was all just there! For almost everything I needed to do the man pages were enough.

When coming from Linux one is not used to this but the OpenBSD and FreeBSD man pages and documentation are actually useful!

Just installed openBSD on my 2007 MBPro (OS X does not update etc.). What a charm.

I bought another MBPro in 2017 because I needed the power and the old one had become unreliable (shutdowns due to heat etc.) I use now the old one for home work and as a remote terminal. Could not be happier.

In case you ever run it from the battery, how is the power consumption?

Never run it from battery: it has become just a backup for fast moves from room to room.

However, I have managed to get the processor get at 1Ghz when not under too much stress automatically (there is a switch which I do not recall, sorry).

The only reason I don’t run a BSD derivative is laziness. It requires more maintenance effort than xubuntu. Back in the distant past I ran NetBSD on SPARC at home.

I’d quite happily be paid to work with it all day but no one wants to hire me to do that. So centos it is.

I tried to use OpenBSD for software development and deployments. It didn't work well because it seemed that packages didn't get as much use judging by inconsistent documentation and sometimes faulty defaults. I didn't care enough to contribute to fix the issues I ran into. Instead I went back to the Ubuntu LTS as before.

I briefly also looked at kFreeBSD with a Debian userland to address the issue but it seemed to be even less well traveled.

I wonder what kind of development and deployments you do and whether FreeBSD might work better. For instance, FreeBSD packages (ports/pkgng) has two branches: latest and quarterly. I've found latest often closely tracking upstream releases.

My work consist of Elixir/Erlang development and a bit of deployments with Kubernetes/Helm/Kapitan/Jsonnet. I primarily work on iPad and login to the FreeBSD workspace via Blink.sh/Mosh and it all worked really well.

That said, if your work involve Haskell and Stack, you might want to stick with FreeBSD 11.2 since Stack still hasn't release a version that supported FreeBSD 12 (which broke due to FreeBSD 12.0 change to inode64)

My motivation for trying *BSD over Linux is for security/stability which is OpenBSD's goal more than FreeBSD.

I always thought BSD is surviving due to its liberal license allowing many companies to use it as a base OS for their prodcits without any liability to open their code leading to financial supports.

From a consumer perspective where a license choice doesn't really matter, I don't find much reason to use BSD over Linux, though I keep a FreeBSD just to keep up with the difference.

I saw someone post that "FreeBSD is a professional operating system for professionals" so consumer perspectives don't really matter. That Netflix, Whatsapp, Sony and most of the world's major internet traffic routers run FreeBSD, via Juniper Networks, is more significant to me.

Sony also deploys FreeBSD in one of their major consumer product: a PlayStation 4, which runs Orbis OS, a fork of FreeBSD 9[1]

[1]: https://www.phoronix.com/scan.php?page=news_item&px=MTM5NDI

I would like to use FreeBSD myself on my laptop, but I really need those intel WiFi drivers to support 802.11ac

Have you tried OpenBSD?

I really like BSD myself, I used it as my primary OS for both desktop and server back in the early 00s.

Nowadays, with a container-centric cloud architecture, it seems like BSD doesn't really have a good answer to Kubernetes, Nomad, etc. If anything, the projects are experimental and/or hobby projects.

Which is a bit bizarre because FeeeBSD jails were world's first containers on a vaguely desktop-ish unix-ish systems.

The crowd just never had the hype power of Linux and no big player took to developing cool things on top.

It just shows you that having a technically good solution is irrelevant if there's no marketing behind it.

> no big player took to developing cool things on top.

Just one data-point, but WhatsApp engineering was adamant that FreeBSD was a key precursor that allowed them to scale 2m+ connections per host on their (rather beefy) boxes: https://youtu.be/TneLO5TdW_M

> It just shows you that having a technically good solution is irrelevant.

True. I've seen many promising projects die/stagnate at the hands of unpopularity, *BSDs wouldn't be the first. Sigh.

It maybe also shows how brainwashed or resume-oriented the tech crowd is.

Network effects are real though.

I remember using Tru64 and HP-UX containers before any BSD got them.

And Solaris zones. But FreeBSD is far more desktop oriented than anything Sun or HP put out (and yes I ran Solaris as a desktop OS for a bit, but still).

Exactly, Docker was able to get a lot of traction because of a variety of reasons (timing and usability, to name a few), even though from a technical perspective it's very much a suboptimal implementation.

And right now you just have a huge network effect around containers + linux: the tooling + support is the most compelling reason to "suck it up" and just fire up Ubuntu and call it a day.

Given its scope, iocage has been an excellent way to manage jails for me. What docker brings to the table is a fairly easy way to build the containers themselves.