Slashdot Mirror
Debian NetBSD
bXTr writes "Interesting project over at SourceForge. Quoting from the website, 'Debian NetBSD is a port of the Debian Operating System to the NetBSD kernel. It is currently in an early stage of development and cannot currently be installed from scratch. Instead, a tarball of the current envionment is available and can be extracted into a handy directory on a NetBSD system.' Check out the reasons why they're doing it and some interesting commentary at DailyDaemonNews on this."
not so much what's wrong with the FreeBSD kernel as what's wrong with linux user land. I would like to see the FreeBSD installer & ports system, to begin with, exist on a linux distro.
I'm new to FreeBSD, so I might be wrong, but it seems like linux gets more of the new 'sexier' things. sambafs was on linux first, numa, IBM s/390 port - those kinds of things. not to mention binary support, which ok, there's linux emulation in FreeBSD, but it's easier to do it straight on linux.
Really, the convergence of Debian package management, GNU utils and NetBSD kernel isn't all that special and WILL NOT create a stronger, unified, easy-to-use UNIX variant.
Please, try Mac OS X; there's every advantage to it without all the traditional UNIX disadvantages.
My hope is that OS X will unify the BSDs into its proper place - at the top of the OS food chain. Many Free/Open/NetBSD users are coming to that conclusion as are many Linux users, beset with flaky kernels and horrible OS packaging.
Apple OS X and the *BSDs will be our answer to WinTel/Linux obsolescence.
If you don't need third party application support or kernel threads, however, FreeBSD has a much more solid, reliable kernel.
It would be excellent if you could maintain different machines with different kernels as needed, but have everything on top of that be Debian (both because Debian is excellent, and because supporting a heterogenous OS environment is a pain best avoided if possible).
I started using linux because it had the hardware support I needed, and support was 100x better. But it wasnt stable enough for my server, so I ran freebsd. But that was a few years ago. Ive always been able to explain to my friends who run BSD, that I need SMP support, so I run linux. But its also how linux has better configuration utilties and drivers. After using linux for years, I know where everything is, easy to setup and fix.
...' - Isaac Asimov (1920 - 1992)
Now, Linux is rock solid, and I get to laugh at my friends who cant X setup on thier freebsd boxes. But then, by the time a good bsd distro will be out, newer and better linux kernels will be out, with new vm's and more features.
-
The most exciting phrase to hear in science, the one that heralds new discoveries, is not 'Eureka!' (I found it!) but 'That's funny
The original proposal was for Debian OpenBSD:
Debian OpenBSD topic
Debian OpenBSD txt
This is oh such a troll !
Giving choice has never been a problem in the Open source world, and the good ol' argument about "all those people are wasting their energy, they should team up !" doesn't take into account the way people work.
If anyone likes a project, he will join it. But if he doesn't, he will start hims. Is this a problem ? There is no such think as a best solution for the environment/distro/kernel choice ; being able to take whatever best fits your need is hence a big plus.
Singularity stupid: stupid gotten so dense that no intellect can escape
RedHat -> Slackware -> FreeBSD -> Debian
For desktop OS I find Debian Linux more convient than FreeBSD for two reasons:
--
Ilya Martynov (http://martynov.org/)
What happened to the concepts of diversity, hybrid vigor, competition, and cooperation?
With Microsoft we get a monoculture.
Are you suggesting the same for all other OSes?
If nothing else this project encourages and explores compatibility issues, source examination, bug catching, performance tuning, and a bunch of other things, if only because a new, fresh, set of eyes (Debian) is looking at old things (BSD), and the other way around, BSD people looking at Debian things.
This cross pollination can have so many surprising and unexpected benefits too. Like the fact that if the kernal is BSD and the userland is Debian... it means you could, besides a little project called Fink, place an entire Debian OS layer on top of Apple's Darwin or Apple's OS X.
Then there is the ports system, which sounds very good to me. It's currently a BSD thing, but there's nothing stopping it from running on top of the Debian-netBSD distro, with work, and therefore stopping it from working on GNU-Debian with just a little more work, with 'work' and 'little more work' being subjective here.
These are just obvious speculations on my part. Many more advantages can be found, I'm sure, of this type of project.
GPL Deconstructed
If I saw any indication that Debian could reduce that chore more effectively than FreeBSD, I'd set aside another three days. At the moment it looks like a toss-up at best, so I stick with what I have.
Well, I can guarantee you that Debian will solve virtually all of your package problems. I love the FreeBSD kernel, but wouldn't think of using any system that doesn't match up to Debian in terms of package management. And none that I know of do, so far.
To be honest, I've tried FreeBSD (really tried),and and really want to like it, but coming from a Debian background, FreeBSD's package management is much too primitive to consider.
Debian not only automatically downloads and installs packages along with their dependancies, it will set up virtuall all programs properly - they run out of the box. Daemons come up as soon as you install them, and reload when upgraded. Upgrading a particular package will upgrade dependant packages. Installing an X app will automatically add it the menus in almost all window managers available. Uninstalls go just as easily. Conflicting packages are managed, so that your favourite pager is known, even with many different paging programs installed. RPMs can be imported into the system, and there are importers for other standard packages like gtk/windowmaker themes (and CPAN too, if I recall correctly).
What else? CDs can be built automatically from the packages on your system (using an add-on package), kernels can be automatically built and patched. Hell, that's enough. You should be convinced by now. In fact, I really can't believe that Debian hasn't replaced RPMs and PKGs by now. I looked up the BSD family's new OpenPKG system with hope, expecting them to have at least included debian's features, and perhaps expanded on them, but no. It's still the same old last generation stuff. Which, frankly, is just crap after using Debian. Did I mention that the latest version of Debian (the current one, in FreeBSD terms) has over 8000 packages available? Gotta love it =)
In short, if package management is a priority for you, then use Debian - it's that simple.
personally, I would like to see a BSD distro with ports and all, but with a linux kernel.
I agree. Although I think the BSD kernel is arguably superior, having a Linux kernel would allow the rapidly increasing number of commercial applications that run on Linux to work.
The big, ugly, problem for me (and almost anyone else that's really worked with and appreciated the real power of *real* Unix, as opposed to Linux) has always been the GNU utilities. They're acceptable, but just barely. GNU Documentation stinks when iut's there at all, at least partly because even most FSF-backers recognize that man pages are the expected form of OS docs and info pages are a hoppeless GNU-ism.
The GNU utilities insist on using their own hopelessly convoluted syntax, (especially the hideous "--" options, another perversity enforced by the gnazis that intentionally creates a gulf between the GNU wasteland and the civilization of the Unix/BSD world.
The BSD utilities are one of the best reasons to run BSD - they are orders of magnitude more stable and standard than their GNU hack counterparts. The code for many of these utilities is indeed old, but has not remained static: The BSD utilities provide a level of maturity that GNU will probably never reach, simply because structure and gols of their organization forces the BSD folks care about such things, while that of GNU seems to ensure that that level of care and attention will not be lavished on the code. In my mind, this is a distinction that is far too often overlooked.
"The future's good and the present is nothing to sneeze at." - Roblimo's last