Post a Comment
I've been following the development of dfly closely, and I'd like to thank every one that has contributed to make this great release happen.
Dfly is made up by a small, low drama, friendly team (I despise the word community) where one can learn a lot of things and have great time. Which is basically what free/open source is about.
Cheers,
Stathis
I agree.
Still, after over five years, the enthusiasm shows. If I stupidly bypass all technical advancements, I believe that DragonFly BSD has genuinely managed to innovate some of the long-standing development and collaboration patterns in the BSD scene. Switching to git is one example, the other could be the cooperation with the NetBSD project on packaging third-party software.
Someone asked the classical question: why? If you are into operating system research, teaching, or development, DragonFly BSD is an excellent choice to study and develop new ideas on. Obviously it is also a production-ready, multi-purpose operating system capable of acting both as a server and as a Unix workstation.
Your question is a bit silly, because a simple Google or Wikipedia query would give you much more information:
http://en.wikipedia.org/wiki/DragonflyBSD
(Still, I voted you up, back to 1, because I feel that such questions should not be modded down. Much people have such questions when they face a new OS.)
shouldn't that kind of info be available on The Dragonfly BSD site?
also.. I understood less than half of the wikipedia page. Symmetric multiprocessing, LWKT ... ports/messaging all gibberish to me, and still gave me no understanding of what DragonflyBSD is or why I might want to use it. Let alone why it might be better/worse than any other BSD distro?
now, you're gonna reply and say "if you don't know what they are Dragonfly isn't for you" and perhaps you'd be right. But a bit of plain English to explain it's pros and cons would be great.
Great! Recently I loved the way DFBSD works. It's not as bloated as FBSD, it doesn't incorporate half-baked stuff [except of hammerfs, but it's understandble] and I had fewer problems with DFBSD, than with FBSD. I had to admit the failure of FBSD - IMO it went a wrong way. Most of the things in FBSD are overcomplicated to me at some point. DFBSD doesn't lack simplicity, so I find it perfect alternative for FBSD. IT actually reminds me of my everyday-use OS of choice, which is OpenBSD.
I'm surprised by how much work such a small team is able to get done. I've tried the DF through Virtual Box and it is quite responsive even in a virtual machine.
The only thing that is stopping me from using it is the lack of nvidia drivers. Without the drivers the fan on my card runs full blast 100% of the time.
I seem to remember that originally a key motivation in forking DFly from FreeBSD was Matt Dillon's (DFly founder and leader developer) disagreement over their SMP strategy. DFly has been engineered to support multiprocessing in a quite different way to most other contemporary OSes. But as far as I could tell from reading the Dragonfly BSD Digest (http://www.shiningsilence.com/dbsdlog/ - a very interesting blog) this work has not yet been completed.
The emphasis seems to have shifted to clustering work etc. It's interesting stuff, to be sure, but I'd still like to see the final results of the SMP strategy.
Their HAMMER filesystem looks very interesting, though. The "high level journaling" stuff looked really cool as well but I don't know to what extent they use that currently.
I'd be quite interested to see a summary of what Dragonfly's benefits are *now* over other OSes.




it's a fork of FreeBSD.