Competition is always a good thing, even in the OSS world.
Linux is just a kernel. The problem is that outside the kernel there is perhaps already too much competition. E.g. there is glibc, eglibc, bionic, dietlibc, etc. There is SysV init, systemd, Upstart, OpenRC, etc. Then we have GNOME, KDE, Unity, Xfce, Cinnamon, and MATE.
For practically any component in a Linux system, except the kernel itself, there is already a lot of competition. The BSDs add four more, albeit incompatible (with Linux and each other), user lands.
I think the FLOSS world would profit from less fragmentation and more focus on making the good projects better. Of course, polishing existing work is not as much fun as writing your own ;).
First, a lot of code migrates around, look at any of the release notes for any BSD and watch how much is from another BSD project. These are not silos.
Don't get me wrong, I love FreeBSD and use it for somethings, but they don't really take security very seriously. Theo just gave this lecture with examples http://tech.yandex.com/events/ruBSD/2013/talks/103/
As a Linux user who hasn't dabbled in the BSDs for a long time, I can't but agree about having them around. My first reaction to this headline was to find a donate link, and I don't even have anything running OpenBSD, here or at work. I'm not sure uniting would be a good idea, however; having separate projects focused on separate needs seems to be a good thing to me.
I disagree that BSD market share is eroding. It may not be gaining as many desktop users, but it's still popular on the server side, especially on file servers due to support for ZFS, is used as the base for other operating systems like MacOS X, Juniper's JunOS, and even Playstation 4's OS. Netflix adopting FreeBSD also means a very large portion of bits being shuffled around the Internet are off FreeBSD servers.
It certainly would seem like merging build farms or continuous integration systems for the ports should be doable. The setup for such a thing wouldn't be free, but it would benefit all the BSDs if they could agree to it, and make this business of supporting esoteric hardware a little less of a sticking point.
The thing is they had their schisms for a reason, and inability to work together is kind of baked in, even if much of the code does in practice cross pollinate further downstream.
The other question is if it's possible to get the benefits of the strange old hardware support without the old school power consumption. I think many relative youngsters would be surprised by how thirsty some of those classic UNIX boxes were, and the hardware must be close to dead now. At some point the only remaining working SPARCstation will be used to build OpenBSD.
Then re-read my comment within the context of this discussion, which is OpenBSD being on the verge of bankruptcy.
If I believed all the BSD forks had a chance to live long and prosper on their own I wouldn't be saying that, but right now it seems to me that none of them have any real long time chance to remain generic OS on par with Linux (except possibly FreeBSD but it's not even certain anymore). Linux just moves too fast these days.
Your comment doesn't make more sense from reading it more often. Saying "because I don't understand how the BSDs work or their development processes, I think they should stop existing to prevent them from having to stop existing" is still nonsensical, no matter how you slice it.
It's nonsensical because you misunderstand it. He is saying he thinks two or more of them ought to combine efforts in order to prevent all of them from failing.
Two of them "combining efforts" is removing one. That is no different than one of them "failing". That is the point. Contrary to armchair OS designer belief, you can't just say "hey guys, stop working on your project and go work on someone else's because I don't know what I am talking about and think this will somehow help something". Go tell linus to start working on mercurial because you think he should. What is the result of that conversation going to be?
> Competition is a good thing, so we should unite bit-parts that are in danger of being stomped out to ensure the dominant player still has competition in the future.
"Unite bit-parts" is stomping them out, that's the point. Even if you just magically convinced all the openbsd developers to stop developing openbsd and go start working on an OS they don't like instead, that would just be getting rid of openbsd to try to prevent losing openbsd. That makes absolutely no sense.
BSD stopped 20 years ago, if you are worried about losing it you are a little late. This weird misconception that because openbsd is begging for money that other projects with the letters "BSD" in them are in some sort of danger is bizarre. Inventing a problem that doesn't exist, then suggesting doing something that wouldn't solve it as a solution is simply insane.
I doubt you were truly unable to determine I was not talking about the Berkeley Software Distribution, but the family it spawned. Which leads me to conclude you are being obtuse. For... pleasure? I don't know.
No, I am pointing out something very obvious and important that you seem to have missed. A bunch of totally independent, distinct projects don't all vanish because one of them wants money. The entire premise that "BSD" would vanish makes no sense.
Right now I feel like linux is slowly eating all the market share, if it continues that way the BSDs will regress back to the lines of Hurd and Plan9.
Competition is always a good thing, even in the OSS world.