The Case for FreeBSD
essdodson writes "Scott Long of FreeBSD release engineering team describes some of the finer points where FreeBSD continues to innovate and display its mature development environment. Items such as netgraph, geom and incredible desktop support by way of Gnome and KDE." From the post: "While I strongly applaud the
accomplishments of the NetBSD team and happily agree that NetBSD 2.0 is
a strong step forward for them, I take a bit of exception to many of
their claims and much of their criticisms of FreeBSD."
I just installed FreeBSD this morning... I must say, straight off the iso, a quick install had me up and running pretty darn fast... much quicker than any linux distro I've tried in the recent past... Now if only I could figure out how to get visual studio to run under it, I could ditch windows... stupid work... stupid requiring development on Windows...
;)
One serious thing about FreeBSD over linux distro's... It feels like it has more of a structure, especially when installing utilities and apps... I find with linux distros, the stuff included feels like it's all over the place, hard to find where things end up installing... but I'm really a vxworks fan... so take what I say with a grain of salt...
---
Programming is like sex... Make one mistake and support it the rest of your life.
I don't see why people are so worried about advocacy. If you're not making money, what is the difference? Continue to refine the thing and get what you want out of it, and if other people don't get it, who loses? Personally I have a use for only a couple of operating systems now, and they are Linux and netbsd. netbsd because it runs on just about everything, and Linux because it's most supported. It's nothing against FreeBSD, which I simply don't need. The point is, I use whatever fits the job and if that was FreeBSD then I'd use that. The best fit is determined partially by functionality and partially by familiarity...
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
In fact, for those who haven't tried it, it's quite an excellent full-featured Unix, with everything you'd find under Linux. In fact, it's fully binary compatible with Linux.
The only difference is that it does things the old way -- vi is vi, not vim, and you get sh, csh or tcsh instead of bloated bash. It doesn't have anyone pushing for "ease of use," though it's about at the level of slackware, except with ports, the greatest package management system known to man. Gentoo's portage doesn't even come close to the flexibility and reliability of ports.
Internally, it runs great, because it's not doing things the kernel shouldn't do to boost benchmarks. It's not deeply involved in corporate America, but remains strong due to good management.
Plus it's far more secure. With how much Linux websites are hacked these days -- see http://zone-h.org/ and check out the statistics section, at least 70-80% of website hacks are Linux based -- I wouldn't run it on Linux. FreeBSD is the obvious choice, as it runs its services flawlessly.
- - - - - Fear not the reaper, but my shiny white teeth.
I also heard that Windows used or at least used some BSD work in it's internet capability push years ago. One question will always dog me: Why aren't the BSD's as popular with their very good license at least in the eyes of the IBMs and HPs?
Also, the development is getting very political, this also scares off people.
Heh, give it time. The contest just started five days ago.
Still, from reading the mailing lists, creating a new logo is pretty unpopular in the FreeBSD community. That unpopularity stems from people thinking that this logo will *replace* Beastie, which isn't the case.
NetBSD 2.0 is a higher-quality release than FreeBSD 5.3 on the IA32 platform. There's just no other way to put it.
My experience with FreeBSD is that the 4.x branch is rock-solid stable, fast, and everything works as it's supposed to.
NetBSD has basically reached that level of quality, with better performance.
FreeBSD 5.x has been unstable for me at best. While the userland programs are pretty much the same, the kernel-level changes have killed reliability. Furthermore, some of the much-touted new features simply do not work yet. I'm sure the SMP performance is much better, but I don't have many SMP machines. I've had problems with hard lockups, just doing things like trying to combine vlan and pf. The bridge interface, afaik, also, still doesn't work with pf.
As far as packages go, ports has more packages, true. Still, rarely has there been something not in pkgsrc that I absolutely needed. Pkgsrc is also much easier to work with, and far more friendly when it comes time to upgrade things. Portupgrade is an abortion, especially compared to even *gack* portage from ricerloonix.
There are reasons there's a buzz around NetBSD these days -- and reasons FreeBSD isn't getting the love it used to. I don't know whether the FreeBSD developers bit off more than they can chew, or if they just are rushing things out the door. But until they get their act together and put out a 5.x-RELEASE that truly is release-quality (by which I mean, all the features *work*, and the drivers are supported the same way), I'm going to be using NetBSD and advising my friends to do the same.
A decent number of them are marked BROKEN.
If by "a decent number of them", you mean "1.5% of them" (192 / 12396 at last count), sure.
Gentoo has superior coverage in portage.
Gentoo may have fewer ports which are marked as BROKEN at any given time; but does it actually have fewer broken ports?
Tarsnap: Online backups for the truly paranoid
The NetBSD team were not criticising FreeBSD: basically, NetBSD stepped up their advocacy as part of NetBSD 2.0 release, including some whitepapers on performance comparision between NetBSD and FreeBSD. If anything, the BSD camps all have decent respect for each other, and honestly, Scott suggested that there was more animosity from the NetBSD camp that I think is the case in reality. All of the BSD camps could do with better advocacy, and Scott's post is more an indication that none of them are doing very good marketing, and as soon as NetBSD stepped up the marketing, the other camps (i.e. FreeBSD) felt they weren't getting a good rap: but really, the issue is, that FreeBSD guys just haven't been out there pushing their case as hard as they should really be.
...due to the heavy trolling I got last week regarding my comment that OS security and usability are 50% admin skill and 50% OS distributor integrity.
I'm learning more and more that OpenBSD definitely needs an admin that is more highly skilled admin than most Windows or Linux admins. I've definitely made progress in my implementing of OpenBSD, but I still say that my axiom holds true (see my SIG): With most OSes, if you have a competent admin, then you can have a secure system. OpenBSD might up the ante with oddball features to ensure security, but until those are implemented in other mainstream systems, they don't apply. Additionally, you really need to have very strong Unix skill to use OpenBSD, so it flies right in the face of my theory. Where most OSes would have the admin skills required at 50% competency, OpenBSD requires something more on the order of 80% competency in order to get a usable box.
-"...bad old ideas look confusingly fresh when they are packaged as technology" - Jaron Lanier (Digital Maoism on Edge.o
I've been using Linux since around '96 something, first Redhat, then Slackware and recently Gentoo when I got my AMD64. I tried FreeBSD for the first time a few months ago when I had an old 200mhz machine that I just wanted to use for something, and since that seemed to work ok (a very basic install, no X or anything like that) I decided to give FreeBSD/AMD64 a try when I had to do a reinstallation due to hardware changes.
I downloaded a minimal boot CD, burned in, booted installed the base system over FTP and then X, KDE etc via ports...
After only a few hours I was totally confused. Everything just worked!! Well, almost everything. I had some problems with the soundcard, that was solved thanks to great documentation pointing me to a very logical solution.
I'm still a bit lightheaded. An operating system just can't be this good, I'm probably going to wake up soon.
I have S3 suspend/resume working on both a Gateway 450ROG and a Toshiba 4600. It was as simple as adding a few lines to /etc/sysctl.conf. I don't have any experience with Linux but the ACPI support in FreeBSD 5-STABLE is what convinced me to upgrade from 4-STABLE.
--Ecks
Time is never free, neither mine nor anyone i know.
I prefer to go the extra mile when setting up a server to know that it will work exactly as i want it to, and for that, FreeBSD is a magnificent platform; you seem to advocate for a prepackaged solution. Who is right ?. If it gets the job done, both of us.
Don't get me wrong, i do appreciate Linux, especcialy lightweight non RPM distros á la Slackware, i will, as time permits, experiment with Solaris 10 as well.
Still, if there is one thing lacking from FreeBSD, it's not Java but a native Oracle port (Not FreeBSD's fault of course)
I feel your pain Mike, but you haven't gone into specifics. It just looks right now like you've not gotten your way about certain directions the core is going, and you've taken your marbles home.
Good luck to ya, I hope you can take your expertise with BSD and make Apple's offering that much better. I just am saying that your post is lacking specifics.
I certainly understand how it might look that way, but it's really not the case.
The BSD community at large are painfully honest. When somebody complains about some missing feature, you usually hear "Yeah, it's too bad we don't have that, you should use something else if it's important to you." Meanwhile, in the Linux world, even with practically the same complaint, you'd hear "You shouldn't be using that, and we don't want it in our system."
Now, that could be because a program is missing certain functionality, because the kernel is missing some feature, etc.
Unfortunately, what I've seen of Linux users bashing BSD, is always uninformed nonsense. I think the most popular one is complaints about the lack of a GUI installer from people who have probably never even used it. As if the BSD installers somehow aren't usable just because they don't have RedHat/SuSE logos for you to look at while your partitions are being formatted.
So, there is a huge difference between the criticisms you hear about BSD from BSDers, and the criticisms you hear about BSD from Linuxers.
Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
Ha! That article is so misinformed. In fact, FreeBSD has the worst most buggy networking stack of all the BSDs due to the buggy "SACK" code which does not properly implement SACK, according to the authors of the SACK spec. FreeBSD 5.3 Release went out with a thoroughly buggy TCP stack. It's hilarious how FreeBSD is touting their SACK code when it's actually something to be ashamed of.
Ditto for the SMP locking they're touting as an achievement. They've been fumbling around for a year and it's still not right.