Are the BSDs dying? Some security researchers think so (2018)
csoonline.comThere's also this other BSD derivative out there called Mac OS: https://stackoverflow.com/questions/3446231/how-closely-are-...
Depending on your point of view this is or isn't BSD. But either way it undeniably includes a lot of BSD variants of tools (like grep, sed, etc.) and BSD licensed code.
Other than that, I've not really encountered BSD in the wild, ever. I know some companies still use it. It's just that I've never crossed paths with such companies in my career over the last two plus decades. I've encountered some companies sticking with Solaris for unclear reasons (pain in the ass to deal with these days) but that's been a few years.
The "it's more secure" argument seems to come up a lot and indeed is a strong value in the BSD community. However, you could legitimately wonder if this is more a case of security through obscurity than a technical reality these days. So few people use BSD these days that hacking it has got to be a pretty specialist skill for a wannabe hacker and probably not worth investing a lot of time in given the limited number of interesting targets. Not necessarily a bad thing if you want to keep hackers out but not exactly a user growth strategy either for any of the BSDs.
The impression I got, and it's only an impression based on various blog posts, articles and mailing lists over the years, is that OpenBSD is the most secure operating system even though it's mostly programmed in C (and that they're not looking to replace that with Rust, et al) with mitigations like W^X and pledge, etc; and the fact that their code base is the smallest. I mean I know this is a wide and complicated area with different classes of bugs and vulnerabilities, but still I thought it was generally accepted.
That would be my impression as well.
Linux does have a much wider user base, however, and that enables the community to even stumble across problems more frequently, while a smaller project like OpenBSD might have to orchestrate specifically-themed hackathons and auditing sprints (as they have). But I would take "security by choice" over "security as a byproduct" any day of the week.
Keeping in mind the breadth of resources that aspiring kernel hackers have access to when introduced to Linux, contrasted to OpenBSD's relative scarcity, it makes the latter quite the underdog success story.
(That is not to say that Linux or the larger of the BSDs is the product of monkeys randomly typing on VT100s -- there is considerable and commendable skill in these projects as well.)
Nope. They are alive, well and kicking arse on multiple fronts.
(2018)
Added. Thanks!