OpenBSD 5.9 has been released a few days early! As always, OpenBSD doesn’t do a very good job of summarising the most important changes in this new release, but that’s okay – OpenBSD isn’t targeted at people like me who know very little about the BSDs. It doesn’t really matter – those of you using OpenBSD were probably already aware of what was coming anyway, and if not, the release notes will still make complete sense to you.
Thom, if you ever do find time to explore the BSDs, I think you should start with OpenBSD. While it has less hardware support than FreeBSD, I’ve found it to be the most accessible of them. If you’ve ever used Slackware it’s very similar, and the installer is quick and painless. While it isn’t perfectly suited for desktop use, it can be tweaked to be desktop friendly, though I feel its strongest suit is in embedded systems and servers.
I moved to OpenBSD from FreeBSD about 6 years ago specifically because FreeBSD had *worse* support for my hardware. Straight Intel stack (laptop, desktop, whatever) with integrated graphics and Intel Wifi, and FreeBSD has had problems since 2008 or so. OpenBSD has supported it all out of the box perfectly.
YMMV of course. I’m also not sure that FreeBSD supports as many hardware platforms as OpenBSD (which is only going to drop VAX support next release). Correct me if I’m wrong.
I was thinking of peripherals such as Nvidia video cards and certain WiFi chipsets, but you’re correct that OpenBSD supports more platforms than FreeBSD. OpenBSD’s hard stance on open source drivers vs. binary blobs means some hardware will never be supported until the manufacturer opens up.
I’ve also found that, in the past, FreeBSD was a bit more responsive on the desktop, but lately OpenBSD has gotten much better in this regard.
I need to find time to get back into using OpenBSD. I loved it when I did use it, especially for servers and building firewall/gateway boxes. It’s about as simple as you can get. Working with Microsoft stacks in my current job, I appreciate even more just how straight-forward OpenBSD is. I wish we used it!
I’ve consolidated my workstation and gaming PC into one box in the interest of saving space in our small house, and if not for that (I have an Nvidia video card) OpenBSD would have a permanent place on this system. I still have my old Core 2 Quad workstation as a backup, and it runs OpenBSD flawlessly, but (again to save space and power) it is now a dedicated Plex and file server.
These days I do most of my OS testing in VirtualBox anyway, as it has become much better at hosting non-Windows OSes. That’s where OpenBSD 5.9 will begin life in my office.
Really cool, but it took me by surprise when I saw the announcement. It’s like, a month earlier than normal.
Broadwell support is great, I have been using -CURRENT for a while but I will move back to -STABLE now that Broadwell support is in a mainline release.
Looking forward to 6.0 on the horizon, with all the great work being done with vmm(4).
What sets openbsd apart is the much greater emphasis on correctness, security and design.
More engineering, less hacking stuff together.
The pride in documentation, the global-view of design and innovation, the genuine innovation and research, the pervasive philosophy … are all traits that benefit OpenBSD.
“Blobs are insecure.”
“Testing on diverse architectures brings out bugs and design flaws”
“Incorrect or missing documentation is a bug”
“Minimal privileges and defense as out of the box defaults”
And so on…
I really believe that properly engineered OpenBSD will eventually overtake the less coherent Linux+distros …
The relatively loose design approach got Linux off to a head start, and rapid adoption. But IMHO those same things will cause it to become harder and harder to sustain. Better design meant a slower start, but guarantees a longer future.
The hare and the rabbit.
Ironically, Linux is becoming more homogeneous with systemd taking over most of the userspace these days. I’m still on the fence about whether that project is good for Linux long-term, or if it’s a cancer that will slowly destroy Linux as we know it. And honestly, maybe that’s not such a bad thing. Linux has gotten too complex, too big for its britches, and no two distros do things the same way, even derivatives of larger distros. In short, it’s a bloody mess.
My dream is for Linux to have a global reset, a fresh start in the vein of OpenBSD where simplicity, proper code, sane defaults, and great documentation guide the direction of the project. I doubt it will ever happen, but a man can dream.
Meanwhile, I’ll be over here running Slackware, Alpine, and OpenBSD, as I’m munching on my popcorn watching the systemd/Linux world burn.
My experience of OpenBSD has been very good, for what it aims to do, and for the reasons you give.
However my agreement is tempered by the knowledge that the NetBSD would take offense at being left out.
My sympathies are with you however, experience has: time and again shown this to be wrong.