OSNews

We all wish things, I wish you hadn't written that stale flame bait. Alas...

Better integrated? They already have their incompatibilies between them and have to "cross-polenize" each other on the points where they lack of innovation/functionality. I love BSD but i disgree when people say that BSDs just create a fork when is needed. They created some forks when could be avoided:

- Matt creater DragonFly because he was tired of people putting "band-aids" on FreeBSD, that should fix at one side and break other. This and the desire of have a better "cluster native OS" created this wonderfull piece of software.
- Bitrig is a OpenBSD less conservative. I still se no other than "lego play" utility to this. I know that OpenBSD guys works hard, but their decisions tend to make all thing difficult, and create this fork.
- PC-BSD = Lack of a stronger desktop initiative by the FreeBSD guys. You know, if they have a installer for "desktop fluffy things" this could be avoided.

Some "BSD decisions" would not change, if Linux didn't existed.

The previous poster meant integration at the system level, not between projects. FreeBSD for example is responsible for the kernel, kernel modules, userland, toolchain, and installer, they are all managed, developed and tested together. Compare to linux where the kernel comes from one place, the userland from another, and they are integrated downstream from the initial developer.

They are as about badly integrated project wise as the 100s of Linux distros, but there are only 4 or 5, so that is an easier problem to deal with.

While I agree with you point in theory, your examples are pretty terrible:


Dragonfly is one of the more distinctive variants of BSD. It has a whole boat load of features not seen in FreeBSD.


Unless things have changed significantly recently, PC-BSD isn't really a fork of FreeBSD, it's more a "distribution". It's point was to give users a no-fuss desktop ready version of FreeBSD. And as PC-BSD is 100% FreeBSD compatible (after all, it /IS/ FreeBSD), I think it deserves it's place as it takes any pressure off the FreeBSD devs from having to cater their limited resources to a multitude of different users expectations (or in layman's terms, FreeBSD can focus on building a solid base and PC-BSD and focus on shipping FreeBSD with the desktop preinstalled and configured to run perfected out-of-the-box).

Yeah. It started with just the desire of make the Operation System more scalable and "cluster oriented from roots", and today we have wonderfull features in it.



Yeah. I´ve used bad example here, sorry. PC-BSD helped a lot with docs. I could use bitrig and mirOS here. Both of them want a more "permissive" or less "Theo centric" version of the ol´good OpenBSD

You assume too much in your statement.

1. One cannot assume that if Linux was never made that the BSD's would take its please in terms of popularly.

One could as easily argue that the BSD's popularly owes its success to Linux, and without Linux the BSD's today would be even less popular.

2. Why would we have better Operating Systems? Having a integrated and userland and kernel/system does not a good operating system make.

It's hard to say. I once read Linus Torvalds saying that if he was aware of the 386BSD project that he wouldn't have started writing the Linux kernel. So all that would be needed to have a BSD parallel universe would be for someone to go back in time and give Mr Torvalds a subscription for Dr. Dobbs Journal. Frankly there is no way to know how things would have turned out. Linux not existing does not mean that BSD would have automatically filled the void. It is very possible that free operating systems would have never been as successful had Linux not been on the scene. Also, without the existence of Linux the outcome of the USL lawsuit may have been different.

And that's a real PITA. It's like releasing a car without wheels. But who cares? the car is there ...

Do your sh#t thoroughly, then release. Never the other way around. Otherwise you're heading to disaster [or Linux].

Except the packages produced by the build system aren't part of the release, while the wheels are part of the car. This isn't semantics; even when extra packages were included on the install disks, they were referred to as "third-party" packages.

I hate car analogies for computers, but a better analogy would be "The car is ready, but it'll be a while before after-market add-ons are available."

Even when packages are available, I find myself building from ports frequently, just for the greater control, and I'm willing to bet that anybody that administers FreeBSD systems professionally will rely on ports much, much more.

Packages build for 9.0 should also work perfectly on 9.1, but the packages build for 9.0 aren't being updated at the moment, either (Though, the ports tree is).

Yes great article indeed! I would like to see more OSnews Originals like this.

Yeah, OpenBSD doesn't get a lot of love.

Good solid OS, lacks features for sure.

Hey!... i give him all my love and two machines (my homeserver and my netbook).

The way I understand it, the package build infrastructure isn't back online yet (at least not as-of the end of January), so there are no packages for 9.1 yet, since it was released after the breach.

Additionally, packages for 9.0 may be outdated compared to their ports version for the same reason.

Thank you, that makes sense. I'll go ahead and continue using the ports tree and compiling. I kind of like it since it feels like the most "FreeBSD" method.

By using NetBSD or OpenBSD instead

RTFM, FreeBSD has a fantastic documentation, follow it.

You just need to install the initial ports tree, and then it is just a simple matter of "make install clean" from the directory of the port you want to install.

Keeping the system up to date is a bit trickier than the software installation process, though. In FreeBSD there are usually n-ways of doing the same (installing ports, upgrading the system, etc) action. Which I think it's what may be confusing you initially.

Edited 2013-02-15 18:51 UTC

Did you just tell me to read the manual to learn how to install software using the ports tree after I said I know how to install software using the ports tree?

You also were asking "what is the best way to install software in FreeBSD," so it seemed obvious to me that your proficiency with the ports system was not that good. Which is why I directed you to RTFM: http://www.freebsd.org/doc/handbook/ports.html

Edited 2013-02-16 03:54 UTC

...which brings me to another question. I'd heard for many years about how strong the documentation is for FreeBSD, and assumed that extended to the handbook as well. But while reading through the handbook, I kind of got the impression that something was just a little... off. I couldn't tell what was current and what was out of date because of my lack of experience wih the OS. But then I got to this part:

"By far, the most popular X11 MP3 player is XMMS." (freebsd.org/doc/handbook/sound-mp3.html)

And literally laughed out loud. Even so, I'm about to install XMMS (first time in about a eight years) just so I can be like "the FreeBSD guys".

Is it appropriate to follow the handbook as the best documentation for installing and configuring FreeBSD 9.1?

Thank you for your guidance!

Lol. Some parts are kinda old, but any technical details should be valid. They do a good job of updating the handbook when things change.

Pity XMMS is down, because they had a little rant about how XMMS is still being used.

Great little amp like the Winamp 2 series before > version 2.81 (they started adding all the useless library features).

Winamp.com still makes 8 million a year last time anybody highlighted it.

Edited 2013-02-18 16:07 UTC

Which BSDs did you try?
Have you tried PC-BSD?

Ultimately though, there's more to an OS than whether your desktop does compositing out of the box.

Edited 2013-02-15 15:26 UTC

If you want something where you just click next, next, next to install and will auto detect hardware and install necessary drivers and media codecs, you can give Pc-BSD a try. It is as user friendly as Linux Mint.

PC-BSD has the same hardware issues as FreeBSD. For example, in order to enable Atheros 9285 wireless support, you have to edit /boot/loader.conf as root and add if_ath_load="YES". This is not something most folks would not feel comfortable doing.

Well, duh! PC-BSD is FreeBSD. Thus, all hardware support for PC-BSD comes from FreeBSD.



ath(4) is part of the GENERIC kernel in FreeBSD 9.1. No loader.conf editing required.

Unless the BSDs get a proper top-to-bottom binary package management working, they will be a OS just for techies and servers. They tried that with pkgng and ... well .... lets see if that works ... has not yet.

I run FreeBSD on several servers for lots of network packet analysis but run Linux and OS X on the laptop/desktop. Severs is great since I do not have to compile and install from ports a lot of packages so no broken ports on the core stuff. But won't trust that on the desktop.

Heres hoping to see FreeBSD on the desktop soon.

2013 should be "the year of binary packages" for FreeBSD.

You can already upgrade the base OS via freebsd-update(8).

And the PKGng project will allow you to do binary upgrades of your 3rd-party software (aka ports) with the ease of Debian's apt-get. All the pieces are in place for 9.0 and 9.1, and PKGng is now the default on 10-CURRENT. All that's missing is for the package-building cluster to be brought back online to build binary packages on a regular basis. The cluster is being rebuilt due to the security incident last fall.

However, one can use ports-mgmt/poudriere to create their own package-building system, and use PKGng to manage everything on their own systems.

That's what we've been working on for some time in MidnightBSD. Our mport package management tools were covered in a BSD Magazine article if you're interested.

I'm planning on doing the 0.4 release with the package tools this year. You can actually use an early version in 0.3 as an option, but it doesn't have all the polish one would expect.

How current is your project in relation to FreeBSD (using them as a touchstone)? I noticed you started the project from FreeBSD 6.x. Are your packages now current with FBSD 9.0, or what? The project looks interesting and I'd like to try it. Will it run in VBox? I've tried past versions of FBSD and I could not get them to boot in VBox...

There is also ghostbsd.org. I installed it once, but the next time, it failed to install. (xfce beta 3 version.) Not a problem here, it was for disk formatting anyway, as I run Freebsd natively.
To more respond to this post, you may wish to try any of the several ways of dual-booting [carefully, following a guide, etc...] That was how I first ran FreeBSD for a year or so. If it had any major problem at any time, I'd just boot into Windows98 and check on how to fix it, configure it, etc.
If you really want to use virtualbox, there is a freebsd.org forum which has probably many posts on the subject. I've read many posts, announcements, but only
speedily as I never expect to have time for virtualization as far as I can ascertain. [I run more than one FreeBSD machine, and it takes all my computing time per se to keep current of the OS developments, aside from normal email, browsing, etc. ]

Sorry, gotta call you on this. Every modern operating system can handle this kind of multitasking with no trouble on decent hardware. Especially with an SSD.

I take it the only OS you've used are FreeBSD and DOS.

I used windows98, and the shareware/freeware plethora not only resulted in BSOD's daily, but the start menu gave scant hint of where many programs could be found, having grown in size.
I've used and could use Linux, howsoever, browsing the forum for one distro daily, I wonder if each poster could not do 20 to 80 percent more fixing/installing/upgrading/developing daily if the poster were running FreeBSD instead. Breakages are posted... for which dual machines... then, often "dd'd sdb rather than sda..." countless times...
My post on FreeBSD's positives were not to put down the ones I consider less efficient, but to point out the advantages of the former, and how it can *indeed* be used as a desktop capable machine, which if I did not make it clear, was the reason for the original post, as
a counterpoint to those who said or implied that FreeBSD would *someday* be ready for the desktop.

The thing that sold me was being able to run pipe commands.

Are you by any chance friends with this guy

https://twitter.com/1990sLinuxUser

And AT&T gave you the source on tape, as they were not allowed to sell binaries.

IOW, it was OSS from the beginning, but not "Free Software".

Just because the source is downloadable doesn't make it open source. It's quite possible that the truecrypt license is not compatible with Dragonfly BSD's goal and policies.
A cursory glance indicates that the license is such that it could not be included as part of any BSD.

Just because the source is downloadable doesn't make it open source. It's quite possible that the truecrypt license is not compatible with Dragonfly BSD's goal and policies.
A cursory glance indicates that the license is such that it could not be included as part of any BSD. "

That's not what "closed source" means. If the article wanted to distinguish between free software and open source, it should have done so, but closed source is unquestionably the wrong one of the three terms to use.

While the license may have change since it was originally rejected it is still not recognized as ether an open source license by the OSI (Open Source Initiative) or as a free software license by the FSF.

http://en.wikipedia.org/wiki/Truecrypt

https://fedoraproject.org/wiki/Forbidden_items?rd=ForbiddenItems#Tru...

Neither of the three terms is particularly good. "Shared source" would be probably the best description of Truecrypt.

I dunno systemd makes my Fedora install boot up nice and fast.