OSNews

My favorite distro is OpenBSD. It's the way ALL OSs should be written, with security in mind. PLUS, it doesn't loose ease of use or potential eye candy as a trade off. You can still install and window/desktop manager and is no easier/difficult to operate than any other BSD (as far as I know).

I always had the impression that OpenBSD was super hard to install and only for the faint of heart. Now, though, having experienced all 3 main BSD's, I find that OpenBSD is fairly easy to setup and administer. It's NetBSD that I find downright daunting. I can setup NetBSD, but then things get pretty wierd for me. No home directory? No adduser command? I guess it is Unix the old school way. I'm not a newbie, but I found it pretty confusing. I have a nice wireless OpenBSD laptop to go with my Debian and Mandriva laptops.

Hmm. I have no experience with OpenBSD, but I do have lots of experience with NetBSD and some with FreeBSD. I don't find either of them terribly difficult to configure or maintain. If anything, both are a lot simpler to setup and maintain than a Debian or RedHat system, at least from the command line.

As for the no home directory thing, the root user is in /root I believe and you can add users with useradd -g users -m -s /bin/ksh foouser, just like you can on most Linux distros. If you specify the -m flag the program will create the /home/foouser directory and copy over any skeleton files. I guess NetBSD doesn't create a /home folder during installation because there are no users that can login (besides root) by default.

Btw, I have a nice wireless NetBSD laptop to go with my Ubuntu desktop. To each his own .

Patience.

I'm sure the BSD-family will receive a lot more attention as PC-BSD and other such desktop versions get better and better.

People cannot stay ignorant of that development for ever.

I do love all BSDs, FreeBSd being my system of choice..

I always wondered what's the main reason why the BSDs don't have the same visibility as Linux.... You always see big companies supporting projects for Linux and not for FreeBSD or his BSDs brothers.

Anyway, I use FreeBSD for anything, it's really great.

I think the reason for the lack of attention on the *BSDs are mainly due to court trial in the beginning of the 90'es.

It came at the worst possible moment for the *BSDs.

The GNU-society (if one can call it that) was looking high and low for a kernel to their system, and while the *BSDs were hampered by lawyers, Linus Torvalds came up with his kernel. And it took off from there.

He was at the right spot at the right moment, and that's why. When the *BSDs came out of the black era it was too late for that train. However, they're coming nicely along now, and I'm sure it's a matter of time. Cross platform applications is becoming increasingly normal, which makes it feasible to have a more varied OS market.

Linus is quoted for saying he wouldn't have written his kernel, if he had known about 386-BSD. So it's not because the *BSDs are bad. They were just unlucky, so now they have to play catch up in regard to userbase.

But be patient. It will happen.

The GNU-society (if one can call it that) was looking high and low for a kernel to their system, and while the *BSDs were hampered by lawyers, Linus Torvalds came up with his kernel. And it took off from there.

He was at the right spot at the right moment, and that's why. When the *BSDs came out of the black era it was too late for that train.

Would RMS really have settled for a bsd licensed kernel? I don't think they would have settled on bsd, linux or not. And the bsds weren't looking for a userland like Linus was.

RMS could always have relicensed it under the GPL.

RMS could always have relicensed it under the GPL.
Err, not without copyright. I suppose it's possible that the bsd folk would have let him but I don't see why they would.

1) all BSD's are performant, portable, and secure.

Don´t want to start yet another flamewar, but I don´t think that any of the BSDs are regarded as excelent performers when compared to most other OSes, with the slight exception of FreeBSD which is highly optimized to the x86 platform, but even then, it can´t touch Linux performance-wise (on the 2.6 series, that is).

But then, I don´t think that performance is everything on an OS experience and I´m looking forward for a time when I´ll have enough spare time to install and play with FreeBSD. From what I´ve been reading about it, I´m sure that I´ll enjoy it a lot.

"it can´t touch Linux performance-wise"

That is a seriously ignorant statement.

Those are pretty big words coming from somebody who's never used Freebsd. I do love how you don't use any facts to back up your statements. Oh and please don't give me performance marks from 1 to 2 years ago. That and I'd like to see performance marks from somebody neutral, not somebody from the linux side of things. Oh and can we not have a flamewar inside another bsd topic please, it's getting rather old. Both Os's are great and should both be used, zealotry won't help either side in recruiting new users.

I tried Crux, Arch, Mandrake/Mandriva, Debian and [K]Ubuntu with various WMs and even custom compiled kernels. Now.. let's compare speed.

Bootup with KDE on Linux ~1 freaking minute.

Bootup with KDE on PCBSD ~20 seconds.

But alright I hear you shouting "bootup speed is nothing, everyone can tweak it blah bla" well.. no the linux kernel boot speed is low. Other services can be turned off but the linux kernel will simply boot slow. You can custom compile but it's still going to be slow.

But for sake of clarity let's forget bootup speed.

General WM (let's use KDE as the typical CPU/Mem hog) performance.. same story. KDE starts faster on PCBSD too. And that's precompiled. So please drop the gentoo argument. General responsivness is great I feel like in a usable system again. And now.. why did I try so many Linux distroes? Mostly because of speed and packaging problems.

I for one am not a big fan of ports. It take ages and sometimes things brake because something cannot be updated during portupgrade -a etc. I think for example that apt-get is much better. But performance-wise I find BSD better. I did some basic tests with compilation speed the difference there was lower but still I saved about 5 seconds from 30 on PCBSD compared with Kubuntu with just KDE running. (PCBSD compiled the stuff in 29 secs, kubuntu about 34 on average from 10 times)

Flame war continues

"Flame war continues "

No.

Have a nice day ,

please continue downloading and using our free software on your kernel and improving it , keep testing our distribution and reporting on them. We apreciate all the help you give us.

- Moulinneuf

But it might be interesting to someone coming to OSNews for the first time and who's never heard of BSD. We're not all well versed.

This is tangental to your question, but in terms of business it's significant.

1) BSD license grants a lot more freedoms to business than the GPL.

2) BSD can't be touched by SCO.

Regarding (2); BSD already had its SCO. It was the AT&T case in the early 1990s.

Regarding (2); BSD already had its SCO. It was the AT&T case in the early 1990s.

That would be my point. If, based on its rights to UNIX, SCO tried to make a play for BSD, it would fail before it started.

1) This is good if they want to use software from other companies, but bad if they want to release source. GPL gives the company more control over the source than BSD. Releasing as BSD would give competitors the upper hand. However, it all depends on what the company wants to accomplish.
2) Neither can GPL and GNU/Linux.

"1) BSD license grants a lot more freedoms to business than the GPL. "

Let's clarify the differences between the two licenses, specially since you used the word freedoms and most people will think of the FSF definition when they read of free software.
The main difference between the GPL and BSD licenses is that if you redistribute software under the GPL you must give the users you're distributing it to the same rights that you received. For example, if you received the software with the freedom to modify the source you must give people the right to do that when you distribute it to others. The BSD license doesn't ask for that; you can receive the software with freedom to modify, redistribute, use for anything, etc, and redistribute it as proprietary software that restricts all these things to the users.

"2) BSD can't be touched by SCO."

This is not true. Any company can claim their copyrighted code ended up in <insert name of free software here>. Like someone mentioned already, BSD already faced that problem.
In the specific case of SCO it's not a threat to anyone because they have no case, to say that you're spreading FUD that only SCO itself or companies like Microsoft would appreciate. Don't put yourself sided with these companies.

Now, *BSD are technically better than GNU/Linux in some ways but you failed to mention any of them. You mentioned only two points, the first one being mostly irrelevant to the readers here (Yay, companies can take the OS that I use and create proprietary versions of it. I'm switching right now!) and the second one is pure FUD that could also be applied to *BSD. You may think you're advocating the use of *BSD but in fact all you're doing is attacking GNU/Linux, which will mainly drive a few GNU/Linux users away from *BSD. Since I do like *BSD I'm sad to see you do this.

By the way, another flavor of *BSD not mentioned in the article is Debian GNU/kFreeBSD, which is the Debian system with the FreeBSD kernel instead of Linux. It's still in development but it's quite stable (well, as far as Debian unstable goes).

Now, *BSD are technically better than GNU/Linux in some ways but you failed to mention any of them.

I mentioned that my post was tangental to the technical discussion, but the points I raise are important overall in the grand scheme of things.

and the second one is pure FUD that could also be applied to *BSD.

FUD?! If I headed a company and we were going to jump ship to a *nix, this would be a factor in considering what OS we chose. But, the popularity of Linux and corporate level technical support would also be a factor for Linux.

You may think you're advocating the use of *BSD but in fact all you're doing is attacking GNU/Linux,

I happen to use Ubuntu as one of my home OSes. Please tell me more about my attacking Linux.

All I pointed out was an aside that BSD has some qualities which make it more friendly to some businesses.


which will mainly drive a few GNU/Linux users away from *BSD. Since I do like *BSD I'm sad to see you do this.

I rather doubt this.

Also, my other home OS, OS X, has its roots in BSD.

security is not everthing to an OS. There are good things in BSD that are not in Linux and good things in Linux thats not in BSD. But asking that question here I feel is just going to start another flame war.

Anyways what I really like about the BSD is that Its development is not as chaotic as that of linux.

Yeah Pike also said "Sometimes when you fill a vacuum, it still sucks." commenting on X windows. Just because he says something doesn't make it true. Alot of people said the world was flat and the earth was the center of the galaxy, but that wasn't true either now was it. I know it may be hard for you to keep your comments to yourself, but if you don't have something constructive to add to the discussion, then please don't post.

although I wish all BSD users well, I would strongly discourage any developer to release code under the BSD license; it's so unrestrictive that it can harm any/all free software projects (itself included).

Licenses should always be chosen wisely. However, there are often good reasons to use a BSD, MIT-like or comparable license. For research projects it is good to get code adopted in the future. We have seen this with TCP/IP, IPv6, X11 and IPsec code.

On the other hand it can be good for companies too. A company can provide a closed source version of their opensource project, without having to worry too much about signing off copyrights, etc. For example, if a company uses a dual-license strategy (GPL and propietary), and some GPL code is imported for which rights are not signed off, the code base becomes contaminated*. Obviously, some companies are not willing to take this route.

* Please interpret this in a positive sense. If a project or company chooses to use the GPL, it is their own right.

although I wish all BSD users well, I would strongly discourage any developer to release code under the BSD license; it's so unrestrictive that it can harm any/all free software projects (itself included).

I hope this doesn't turn into a flamewar but I just want to give two counter examples to your assertion.

First, people release code under the bsd license under the understanding that they won't rely on contributions coming back from the community. It's a matter of whether you think you need licensing conditions to keep your project alive or if it can do it on it's own merits. If you're satisfied with what you can write on your own, then the gpl isn't really needed (in the sense that you implied).

Second, a contemporary example of a more restrictive license harming an open source project. As you may or may not know, apache started a project called harmony last year to make a Free jdk. Since lots of good work has already been done in this department, the first step was to look for stuff to reuse. It was determined that Classpath was the best class library implementation and that reimplementing it would be a waste. However, the final project has to be under a license acceptable to the apache foundation and classpath is lgpl, which is too restrictive. The issue hasn't settled yet, but the licenses are definitely a huge barrier to co-operation. If classpath had been bsd licensed, there would have been no problem.

Now certainly, that can swing both ways but I think that, for the most part, when you declare a licensing goal for a project you're dictating a maximum restrictiveness that you'll tolerate. Less restrictive stuff generally isn't a problem.

My personal preference is that if I was writing stuff and releasing it, I want it to be as easy as possible for all open source projects to my stuff and that corporate leeching is a secondary concern.

Except that it's just not true anymore with several FreeBSD desktop distributions more or less incompatible with FreeBSD.

Look at the amount of *BSDs and remember that Linux distributions are just as divided as the *BSDs.

The situation known from GNU/Linux is also true for *BSDs, though to a lesser extent (at the moment).

This will become an ever increasing problem in the future for the *BSDs, as it has always been on all other platforms ever known to mankind.

Again, drivers and hardware support are the two critical areas holding the BSD's back from adoption--not to mention Linux mind share.

Actually, it is sometimes hardware support that drives me to the BSD's. My main hobby is older laptops (P1 and PII), usually with wireless cards, and sometimes no bootable CD drives. Sometimes debian (one of the few distos to still support floppy-based network installs) chokes on the wireless card as a network install device. For these situations, I'll give one of the BSD's a shot. I have found OpenBSD especially to do well with these devices, especially the wireless cards. YMMV.

Last time I checked the Intel D915 was supported just fine. Now the onboard audio might not work, tell me what chipset it's using and I'll take a look.

Edited 2006-01-23 16:43

The onboard audio is Intel High Definition Audio (HDA) on Realtek ALC860 chipset. Thanks for the info (and in advance, too)!

I hate to give you bad news, but from everything I can find, it looks like the Realtek chipset isn't supported at the moment.