The second release candidate of the 5.2.1 is now available (mirrors). It includes fixes to mksnap_ffs, SysV Shared Memory, devfs, IPSec, GEOM, the ata driver and many others. This is probably users’ last chance for bugfinding as the final release is expected to be less than a week away.
First let me start by saying that I think FreeBSD is a great OS. I have been using it since the initial 4 release. BUT, I am concerned about the slow pace of development. It seems, and I could be mistaken, that the 5.x branch is almost a year behind where it was proposed to be.
Am I just mistaken? If not, what could be contributing to this slow down in pace. Is Linux stealing some of FreeBSD’s thunder?
” Is Linux stealing some of FreeBSD’s thunder?”
I doubt it. Its just that BSD employs a much more rigid devolpment sytle. “Slow and steady wins the race” basically.
Personally I like to try out bleeding edge software so im a big fan of Gentoo.
Preface: I have used FreeBSD personally and professionally for nearly eight years.
There was a time when FreeBSD was unequivocally superior to linux…which is to say there was a time when linux was a joke. FreeBSD was the best choice for a unix-like OS on x86 for most of the 90s.
Times have changed. FreeBSD has lost key developers. The 5.x release has been problematic and reminiscent of the 3.x fiasco. The linux 2.6 kernel is technically as good as anything the BSDs have to offer, and it also HAS A FUTURE. Don’t quote me Apple references – their involvement in FreeBSD is tangential at best.
All of the open source projects are targetting linux. RPMs are becoming the de facto distribution format. Meanwhile the FreeBSD ports tree continues to have gaping holes and numerous broken packages. Don’t tell me to “mail the maintainer!” – If the maintainer does not even know their own port is broken, something is amiss folks. Lets not even get into commercial support – FreeBSD has none, so there isn’t much to say regarding ISVs.
FreeBSD, you were great in your time, but that time is over.
I don’t wanna sound as negative as well — I love FreeBSD — but what “blah” said also represents my opinion of FreeBSD *today*. The FreeBSD team need to get more commercial development support, need to re-examine their goals and where the market is going, if they wanna stay players in the OS game.
“The FreeBSD team need to get more commercial development support,”
That is one of the silliest things I’ve heard in a while. They’ve done quite well with little commercial support. More certainly couldn’t hurt, but quite frankly I think you’re mad for thinking it’s required.
“need to re-examine their goals and where the market is going, if they wanna stay players in the OS game.”
I certainly agree that they need to re-examine their goals, but you have to remember that as a free OS project, they were never in it for the market. That is not to say that they shouldn’t keep their eye on things, but your arguments are too business oriented.
I do like FreeBSD quite a bit, and hope that they are still around for years to come. I dislike some of the crap thats going on currently, but I’m sure that it will all work out for the best. At any rate, even if they do go belly up (unlikely) DragonFly is looking pretty darned interesting at the moment, and I’m sure that there will be users (yes, even business ones) despite the lack of commercial support.
>That is one of the silliest things I’ve heard in a while.
I don’t think so.
>They’ve done quite well with little commercial support.
Yes, but Linux showed us that they did much better when IBM/RH/SuSE/etc got into play and helped with development. Having developers working 24/7 for food, yeah, it works a bit tad better in the long run.
And please keep the right subject when replying.
What’s dying? Development stagnation? 0% market share? At least it’s different than with binary projects because anyone who wishes can pick up development.
The BSD’s have their place. OpenBSD is somewhat innovative regarding security (try PF, try IPT, and tell me what’s more easy). NetBSD runs on a huge number of platforms Linux doesn’t. FreeBSD 5.x also aims more on grid which is why DragonFly was founded, iirc. Plus, there’s Apple too.
When GPL software becomes more and more mainstream, i hope the BSD projects keep existing and actually the code as BSDL will keep existing. They’re currently a base for a company (or rather group of individuals in general) to start competing in a different way than GPL software does. If a company wishes, they can do it even much more propiratary than Apple does because the fact Apple gives back and keeps Darwin open isn’t mandatory. Because of this reason, i doubt BSDL software will totally be abandoned, though also a lot can change in the IT sector and a lot can influence it. Imagine C is gonna be obselete in 10 years.
I sincerely hope the Internet Humanity (or humanity in general) by the time the current monoculture is overthrown has learned software monoculture (or huge majority dependance on a minority) is a Bad Thing.
Torvalds for one always said he wanted a healty competition. 95% running Linux doesn’t fall under that IMO. (Guess he likes this site too
FreeBSD is dying because longtime users are moving to linux. That is how I define dying.
OpenBSD has a niche, there is no disputing that, but it is a microniche.
Does NetBSD run on more platforms? Has it been ported to a wristwatch and a Mainframe? Or does it just run on old crappy architectures no one cares about anymore? NetBSD has always been the most hapless of the BSDs and I question if there are even five thousand independent users. Once again, I don’t even think your claim of NetBSD running on more platforms *available now* is even true. And no there isn’t a huge market for vintage Apollo hardware so don’t tell me about a huge niche being met.
Apple is pushing Darwin. Their key developer used to run FreeBSD but jumped ship. To this point some of the departures from FreeBSD have been acrimonious.
“FreeBSD 5.x also aims more on grid”
?????
“FreeBSD is dying because longtime users are moving to linux.”
Prove that.
Which ports are broken? Some were broken due to the change from libkse being renamed to libpthread. Other than that, the ports tree has been peachy.
Of course the 5.x releases have been problematic so far. They are for testing purposes, not for production use. Unless you’re prepared to deal with problems, you shouldn’t be using the 5.x releases.
None of the developers that left (or kicked out) were key. All the contributers are more than capable.
Jordon Hubbard didn’t ‘jump ship’. He was hired by Apple due to his long time development history on FreeBSD. He can’t do full time development at Apple and also continue to be the core team.
“FreeBSD is dying because longtime users are moving to linux. That is how I define dying.”
Ok fine, that makes argumenting easier. Earlier you claimed BSD instead of FreeBSD, in your title. Your title is wrong, or you mean BSD instead?
I have readed no indications which back this up. What i have readed, is Fefe’s scalability benchmarks which speak before Linux 2.6 but aren’t negative for FreeBSD either, while on the other side of the story an article popped up that on some top10, 7/10 ran FreeBSD.
“OpenBSD has a niche, there is no disputing that, but it is a microniche.”
And, “OpenBSD is dying because longtime users are moving to Linux”?
“Or does it just run on old crappy architectures no one cares about anymore?”
At least it’s used on these according to you “old crappy” architectures. Is “NetBSD is dying because longtime users are moving to Linux.” regarding this?
“I don’t even think your claim of NetBSD running on more platforms *available now* is even true.”
In vanilla 2.6.0 source:
$ ls /usr/src/linux/arch | wc -l
20
After a quick count here
http://www.netbsd.org/Ports
~55
Finally some developers and users don’t like the GPL. They find it too restrictive. Wether they’re right or not, do you think their opinion is gonna change Soon, them devving/using Linux instead? If so why would they change their opinion?
What problem is their with ports breakage?
I’ve rarely had problems and that was becuase I track the ports and keep stuff up todate and ran into a few problems with a few things when the defualt gettext was being changed and that was everything was working nicely with in a week.
People don’t switch from FreeBSD to linux, it’s the other way around
Spiceymikey is right
Ive known many people that have switched from Linux to FreeBSD,myself included. Every day i thank god FreeBSD saved me from the Linux nightmare.
At the end of the day, what really matters is what big business supports. *BSD has been there for a while but never really got the kind of support, as well as publicity as Linux is having right now.
Hopefully the industry is getting away from “what big business supports” and moves more towards “what standards support”. If not, it only shows that we still have much to learn as people when it comes to computing.
Commercial Vendors
http://www.freebsd.org/commercial/commercial.html
Ports are broken? O’Please, how about using the ports system and then tell me about.
Has anyone subscribed to the FreeBSD mailing list. Didnt anyone see that Oracle has taken a look at porting some of their software.
Dont get me wrong, I like linux and yes linux gets more hardware support; however, *BSD has better support.
“FreeBSD is dying because longtime users are moving to linux.”
As Kingston said “Prove It”. I have seen long time linux users moving over to *BSD. How about searching the mailing lists. It seems that its the otherway around. It quite funny to see Linux users passing their own FUD along. Perhaps MS has given you all some good lessons.
According to the computer press, dying is what all types of Unix has bin doing for the last 30 years. And I expect it to continue dying for at least the next 30 years to come.
“Commercial Vendors
http://www.freebsd.org/commercial/commercial.html“
True. Also true for Net and Open. Links available on their respective mainpage.
When (s)he was talking about support (s)he meant “support” as in code and moral supprt. For example, what SGI, IBM, HP do for the Linux kernel. At least that’s what i understood from it.
Regarding FreeBSD 5.x it is widely known some Ports currently do not compile. Freshports has some info on this too. I can give you 2 examples which i find out in my own effort (5.2): security/sfs and security/vncrypt
Ever since Linux reached 2.0, its rabid fans have been going nuts declaring that FreeBSD or the rest of the BSDs, for that matter, is dead. The BSDs never ever wanted to go commercial or mainstream and neither do they aspire to do so today. Just because the project needs to work extra hard to stamp out some bugs in the *TESTING RELEASE* does not mean FreeBSD is about to die!
Every time there is a topic about FreeBSD on OSnews.com, the comments section is full of remarks from people who think FreeBSD is almost dead and that Linux is better. Heck, if you think so, you’ve already moved to Linux! So what difference does it make to you? I, on the other hand, moved to FreeBSD after using Redhat, Slackware and a plathora of other loosely attached *steaming piles of SCRIPTS*! And I continue to try them out today, to see if the “Distros” have changed, and I’m sorry to say, they haven’t!
So guys, give it a rest, FreeBSD is not dying, and if it is, it should not matter to Linux users.
For me FreeBSD can currently offer a better desktop solution than any Linux distro. This is partly due to the powerful and stable base system but mainly to the ports system that allows me to easily install and upgrade the latest and greatest free open source desktop software. Still, I would advise MS Windows users who consider trying out a free Unix-like OS for workstation/desktop purposes to begin with Linux distros like Fedora or Mandrake rather than with FreeBSD.
IMO, FreeBSD could gain more users and perhaps also better commercial support by making it easier for newcomers to adopt the OS. The idea that the base system is developed somewhat separately from the ports system seems good to me, but perhaps FreeBSD could assemble a small ‘desktop team’ from developers who consider it worth while to make FreeBSD easier for wider user base. This ‘desktop team’ could concentrate on porting and developing ‘user-friendly’ free open source solutions, like GUIs for configuring XFree and GUIs for other basic system configuration tasks.
I would like to see this kind of development in FreeBSD but, like I’ve already implied, it should be done seperately from the development of the base system. I quite like the FreeBSD idea that the OS itself is developed as a solid and integrated ‘base module’ on top of which users can add different applications according to their specific needs.
In general, I don’t see Linux and BSD eating each other’s resources and support. As long as the general tendency for Unix-like operating systems is that the user base is growing and hardware support is getting better, what is good for Linux will be also good for BSD and vice versa. Of course, if the general tendency would be in the opposite direction, the situation might be quite different.
Anyone else find it funny that on OSNews *BSD is dying, but BeOS is alive and kicking? Give me a break. *BSD is very much alive. Also, many of the people posting in this thread seem to be under the impression that you have to use one or the other. Personally, I use a combination of Linux, FreeBSD, and OpenBSD. Each has their advantages and disadvantages.
FreeBSD is very refined and contains a much more mature and well thought out userland than the GNU userland found on Linux distros. Also, anyone who has ever attempted to compile a kernel on both platfroms will tell you that *BSD kernel compiles are SOOOO much easier. System upgrades through cvsup are simple and flawless, the documentation is EXCELLENT and the ports system is absolutely amazing. I’ve only encountered a handful of broken ports and these were fixed within 2 days in each instance. I do recommend installing portupgrade though, since it makes the already great ports system even nicer.
However, hardware support on FreeBSD is pretty bad compared to Linux, and FreeBSD definitely isn’t for the point and click to setup your network types. Sysinstall is the closest thing you have to a centralized easy to use configuration tool.
Additionally, Linux is definitely a better multimedia desktop experience, I don’t think even the most serious *BSD zealot will argue with me there.
If you need security, OpenBSD can’t be beat. It’s a very clean and well designed *nix that just *feels* secure. It inherits many of the nice properties of FreeBSD as well, such as easy upgrades through cvsup, the ports system, and excellent documentation.
*BSD doesn’t have every latest greatest kitchen sink feature like Linux because they take the time to design things right. That’s not a slam against Linux, just a simple way of putting the difference in attitude represented by these two families of OS.
Ok, maybe its not dying, but its a matter of momentum. I am sure there were days when UNIX diehards were going on about how Windows was DOA. But here we are and many of them are now irrelevant.
Linux has much more momentum than BSD and is growing in leaps and bounds. A lot of innovation is going into BSD as well.
As for the guy who says BSD makes a better desktop, its not about what a desktop is capable of that makes it better or worse, but more about how easy it is to make it reach its potential usefulness. For example, using Fedora, I only hand configure fstab mostly, (I wish Redhat had a mounting program like Mandrake and SUSE), and the rest is pretty much done for me. I do not have to know what my hardware is or anything like that. Las time I checked, some very popular hardware like Soundblaster Live did not have very good drivers for BSD.
Also, last I checked, Linux was more scalable to more CPUs than the BSDs. The BSDs barely made it past 2 CPUs, but Linux scaled to 8 at the time, and now to many more.
Why so many Linux users want BSD to die?? Let’s just say if it’s died now, how would that benefits to Linux??
It has Sound Blaster Live support… and yes FBSD works rather nicely with popular hardware. I have a nice bit of it laying around and have never had a problem with any of it here.
And FBSD will handle more than 2 procs nicely and there has been massive rework of the SMP code and the related.
I use FBSD for desktop becuase of it’s ease of use… no need to constantly be having to fix stuff like on linux… granted it is not for the totally briandead, but the basic unix commands can easily needed to take care of nearly any thing in 15 minutes. Any one who does not feel that taking a bit of time to learn the OS they are going to be using has serious problems.
>And FBSD will handle more than 2 procs nicely and there has been massive rework of the SMP code and the related.
As of kernel 2.6 – this is not a problem. And actually, the 5.x branch of FreeBSD is using the same scheduler that the 2.6 kernel of linux is. Linux implimented the 0(1) scheduler first and it’s in production now; whereas in FreeBSD it won’t be the default scheduler until 5.3 and that’s simply to help find bugs quicker.
Actually I am willing to argue FBSD makes a great multimedia desktop =]
I use it for all my video and gfx works and never have had a problem and don’t see any reason to switch to another OS.
“Linux has much more momentum than BSD and is growing in leaps and bounds.”
From where I’m sitting, both Linux and BSD are growing by “leaps and bounds.” If you look at many of the basic features in Linux and FreeBSD for example, you see the same new features being implemented in both. Better kernel assisted userland threading, better SMP support, new modular security features, more drivers than neither have ever had before, some manner of virtualized environments (UML in Linux and Jails in FreeBSD (and yes I’m aware of the technical differences thanks you)), more supported thrid party applications, both are working to have fully preemptibe kernels, more platform support and of course, both have their fair share of growing pains.
A number of these features I’ve seen first in FreeBSD, and a number of these features I’ve seen in Linux. Except for a few minor technologies that are only useful in extreme situations, these two operating systems seems to keep up quite nicely qith each other, despite all of the companies trying to take Linux in their own desired directions, and despite the fact that FreeBSD has much less comercial support.
This BSD is dying thing was old before I ever even knew about BSD.
FreeBSD has some very nice desktops; in addition to the latest Kde and Gnome desktops there are others like Fluxbox and Xfce4. I use FreeBSD 4.9 as my desktop and it works quite well. The CUPS printing works well, also, as does USB scanning. I can play *wmv or mpeg files with Mplayer here too.
You can get OpenOffice 1.1 or the commercial TextMaker word processor, for FreeBSD. Both do a very good job. The latest version of Gnome Gnumeric (1.2.6) is also available for FreeBSD.
Gaim? That was just the last one I tried, and that was with a brand new checked out port.
That is but one. And please don’t tell me because I’m a dumb newbie, I have been compiling BSD from source most likely since before you ever heard of it (eight years).
Broken ports have always been a problem.
FreshPorts http://www.freshports.org/ currently reports 58 broken ports, including japanese gaim. Total number of ports listed in FreshPorts is 10359.
Or another way to put it, less than 1% of the ports are broken. 58 / 10359 = 0.00559…
FreshPorts http://www.freshports.org/ currently reports 58 broken ports, including japanese gaim. Total number of ports listed in FreshPorts is 10359.
Ports marked BROKEN are a small problem. The more insidious problem is those ports not marked BROKEN which are in fact broken 🙁
e.g. http://www.freebsd.org/cgi/query-pr.cgi?pr=52859
This problem means that anyone using FreeBSD as a PDC does not allow users to change their passwords via Ctrl-Alt-Del from a windows workstation logged into the domain. There is a patch, but it has been sitting for almost nine months.
There are numerous other similar issues that I have seen over the last year that seem to indicate a slowing development in FreeBSD.
The one big advantage I used to think FreeBSD had over Linux when used for a workstation was interactive performance under heavy load – I had always found Linux interactive performance when thrashing the machine poor whilst FreeBSD remained responsive. I must say that my recent foray’s into Linux have changed that.
Couple that with the fact that common hardware seems to work better in Linux (SB Live drivers took almost a year to be accepted into mainstream kernels, my bog standard Netgear NICs works great under Windows, Linux but have appalling performance under FreeBSD) and it seems that Linux is an all-round easier choice these days.
Just correcting myself …
SB Audigy drivers took a year … not SB Live drivers
The more insidious problem is those ports not marked BROKEN which are in fact broken 🙁
Fortunately, the number is not too large. You can see the current build reports on http://bento.freebsd.org/ Some fail just from conflicts between ports. Personally, I have not run into a broken port (for me) in a long time.
Actually, friends of mine have a more difficult time installing software (finding the RPM’s usually) than I do on FreeBSD.
Couple that with the fact that common hardware seems to work better in Linux (SB Live drivers took almost a year to be accepted into mainstream kernels, my bog standard Netgear NICs works great under Windows, Linux but have appalling performance under FreeBSD) and it seems that Linux is an all-round easier choice these days.
That is odd. I have used several Netgear NIC’s (on several machines) under FreeBSD with great performance. In fact, that is my favorite brand of NIC’s.
My desktop/workstation:
FreeBSD 4-STABLE
Ti4200 (nVidia driver)
Turtle Beach Santa Cruz
82c169 PNIC 10/100BaseTX (Netgear)
Athlon XP 2100
512MB ECC
I even run Neverwinter Nights on it.
RPMs are becoming the de facto distribution format.
For what? The source is nearly always in TAR format with a few in ZIP. Mostly GZIP and sometimes BZIP2 for compression. RPM is barely there. It is usually in addition to the TAR file. I have not seen a single application only giving source out in RPM format.
If you were speaking about the binary format, I would have to say who cares. FreeBSD users mostly install from the source even if a FreeBSD package is offered along side the source.
Meanwhile the FreeBSD ports tree continues to have gaping holes and numerous broken packages. Don’t tell me to “mail the maintainer!” – If the maintainer does not even know their own port is broken, something is amiss folks.
If something is broken just for you, why should the maintainer know about it? Psychic? I wonder how many people just expect Linux developers to just know something is not working for an individual or group of individuals. Some broken packages are due to changes in core pieces that the maintainer may not have updated on his/her system.
Lets not even get into commercial support – FreeBSD has none
I guess nVidia (http://www.nvidia.com/object/freebsd_1.0-4365.html) was not aware of that.
Let’s not forget the fact that FreeBSD is the only free/open source OS that currently has excellent support for AMD64 (insofar as a NTR can offer excelent support
Any of the Linux distro’s I’ve tried have not been nearly as nice (if they worked at all) as FreeBSD on my Opteron.
I’m always baffled why freebsd threads on osnews always boil down to arguments over whether or not its dying and why N subject is a sign of some impending demise. Speaking in terms of the most recent transitions, what we’re seeing are growing pains: freebsd dev has redesigned or is in the process of redesigning *major* portions of their underlying codebase as well as adding a host of new features to the OS. Of course things are going to break and in a lot of cases, unexpectedly. That doesn’t mean development is slowing, in fact, if some of the posters here had taken time to review the commits, you’d see the exact opposite.
Who could possibly claim to be an open-source advocate, yet jump at the chance to discredit the hard work of some of the communities best coders? This is not an issue of whether Linux or *BSD is more stable, provides better performance, or whatever thousand reasons you’ll find to argue over – FreeBSD is one of many open-source projects that deserves at *least* the proper attention to run the code and report back constructively. As a long time osnews reader and *nix user, I encourage most of you to spend less time with abstract arguments and blanket statements more time on supporting whatever OS/codebase/project that you *do* like. BSD development helps progress OSS, you’re only hurting yourselves when you claim otherwise. Lighten up.
If freebsd is dying, how come it works so good here on my 600E laptop? What works:
Winmodem, USB printing and scanning, WiFi, Mplayer (*.wmv files too), sound works, apm works (but has some minor issues).
Some of the installed programs that work (a few of the 355 ports I have here):
OpenOffice 1.1
mysql-4.0.18 and MysqlCC (a GUI frontend to MySQL)
Gimp 2.0 pre and TGIF
Xfce-4.0.3, Gnome-2.4.2
Gnumeric 1.2.6, Scilab 2.7
Mozilla 1.6, with java
This is a fine OS
I guess nVidia (http://www.nvidia.com/object/freebsd_1.0-4365.html) was not aware of that. .
Its really third tier. Linux is on the 5000 series drivers. Linux went past 4365 a long time ago.
“Linux went past 4365 a long time ago.”
Linux did nothing in fact. nVidia is no different than other commercial entities in that it is slow to getting quality updated prouducts out the door. That is the reason that the FreeBSD drivers are behind those of Linux; it has nothing to do with the quality of FreeBSD. The inability of nVidia to adapt quickly has little effect on the quality or utility of major free OS projects.
It should be a near mindless operation for them to rebuild their drivers dayly to support operating systems like Linux and FreeBSD, but they do not. They are not as agile as large unpaid cooperative groups, like those that actually do most of the work on large FOSS projects.
Sure their paid engineers could kick our communities’ collective asses when it comes to designing and producing hardware (it takes a real math skills to do that properly), but then that was never in question, nor is it what this argument is really about.
Nice FUD broken ports people. This weekend I finished upgrading 169 ports, and I found only 2 broken. Now try that in gentoo. A gentoo puter sits next to my desk, and I have been helping out my linux-newbie friend quite a lot. Oh, by the way … How do you remove packages recursively (or Upward recursively, like in pkg_deinstall)?: http://www.gsp.com/cgi-bin/man.cgi?section=1&topic=pkg_deinstall
One reason that FreeBSD development is moving slower is that they try to implement something SUN couldn’t and Linux didn’t: M:N threading (libkse) – and the difficulty of that task slowed down the project. In fact, they have two threading libraries besides the old one: 1:1 – libthr, which is similar to what linux has, and M:N – libkse, which is finished and working, it just needs more testing.
RPM as the standard? lol. So you say debian will switch, yes? OTOH FreeBSD can use RPM. In fact it uses it. Did you ever see how linux-base is installed? They are (in my case) standard RedHat 8 RPM packages But one can use .deb if he or she wants to as well
People do switch from Linux to FreeBSD. I did. Went from Windows/Linux (80%/20%) user to FreeBSD/Windows (90%/10%). Now it’s my primary desktop and I love it.
I’m so much more productive because I never end up doing wild goose chases for some silly little problem. Sysadmin in general a breeze – read the docs, follow instructions, works as expected. I use all my time to do real work.
I used to fear typing ‘make’, now I do that with a smile on my face. Over the last two years I’ve hit 2 or maybe 3 broken ports (that didn’t compile), for all of which googling provided information within 2 mins.
The Linux world seemed like a big hodgepodge that I could never grasp. I felt that I had gaps in my knowledge that I had no means of patching. The feeling of full control that I was seeking always eluded me. Some cutsie gui application would try to do the job, mostly work but sometimes fail leaving me totally lost. And the filesystem hierarchy sometimes reminded me of Windows. Then I decided to give FreeBSD a try and there was no turning back.
I know exactly where is what on my FreeBSD system because it has been very simple to build this knowledge from scratch (handbook and other quality documentation). My system does exactly as it’s told – after the base install I didn’t even have telnetd on my system (after two years I still don’t, it’s a desktop).
To summarize, as a Windows user I thought I was looking for Linux, later I found I was really looking for FreeBSD.
“To summarize, as a Windows user I thought I was looking for Linux, later I found I was really looking for FreeBSD.”
Heh. I know exactly how you feel. Windows is certainly much better now than when I used to use it exclusively, but back in the day, it was bad enough for me to actively go out looking for alternatives. Hype brought me to Linux. It was in many ways better than what I had left behind, but there were many things that bugged me.
The man pages were horribly outdated and just plain sucked. GNU utilities with their ridiculously long options damned near gave me carpal tunnel. The fact that the filesystem layout seemd to change with not only each distribution, but wth each release of any given distribution. The horrible backwards compatibility with older binary only (commercial) apps. Poor performance, needing to be a genius to compile the damned Linux kernel. RPM hell. Terrible configuration issues. Multiple CD installation sets. Linux doesn’t even have a kernel debugger for crying out loud!
In short, I had dug myself out of one hole, only to have fallen into another.
Then I discovered FreeBSD. Great backwards compatibility with previous versions. Ports and packages worked far more often for me, with far fewer headaches than did anything in the Linux world. Compiling a kernel is so easy that a child could be trained to do it in under an hour. The docmentation is second to none, unrivaled, anywhere. The filesystem layout is consistent, and (as far as UNIX-like OSs go intuitive. Same for configuration. One CD for all my needs.
GNU and Linux systems in comparison are a joke, the kind that’s about as funny as a racial slur. The fans are rabbid, uncompromizing, and it’s almost sickening that the big companies are putting so much time and money into it when all that somes out are half baked features that rarely ever work fully or right.
The FreeBSD core team is making some mistakes for sure, but to hear the constant droning of the Linux fanatics and the endless posts about BSD dying is mind-numbingly painful to people who have half a clue.
Some people may still encounter ‘broken’ ports that complain about either wrong version or altogether missing libintl.so. This is due to recently upgraded gettext port. Fortunately there’s an easy fix – install portupgrade and upgrade gettext properly:
# portupgrade -rf gettext -m BATCH=3Dyes
After upgrading gettext you’ll notice that those ports that complained about libintl.so are no longer broken. 🙂
I’ve tried the Linux kernel 2.6.x branch on both Slackware and Debian (I optimized it for my hardware) and found FreeBSD (not optimized) to be much faster than any distro with any Linux kernel branch. I also have had more problems running Debian stable and unstable than using the FreeBSD 4.x and 5.x branches. BTW, there are also big companies and important open source projects using FreeBSD such as Yahoo!, Sony, Apache, …
# portupgrade -rf gettext -m BATCH=3Dyes
This should be:
# portupgrade -rf gettext -m BATCH=yes
Let’s make that:
# portupgrade -rf -m BATCH=yes gettext
and we’ve probably hit the winning formula. 😉