FreeBSD 5.3-BETA5 Available

Nirbo writes "FreeBSD 5.3-BETA5 is now available! Get it while it's hot! Here is the mailing list post. Remember folks, this is currently the last beta that will be released for 5.3, we're only a week from a Release Candidate, and two weeks from a release!"

Bad link

[christopherfinke]

The link in the write-up is bad; here's one that works [freebsd.org].

Beta 5

[whyne]

"Remember folks, this is currently the last beta that will be released for 5.3" Not quite accurate "We will add at least one more beta (BETA6)to accomodate testing."

Re:Beta 5

[Coneasfast]

Not quite accurate "We will add at least one more beta (BETA6)to accomodate testing."

where did you get this info from?

the schedule [freebsd.org] only shows up to beta5, then a release candidate.

Re:Beta 5

[whyne]

the mailing list post [freebsd.org]

Anyone knows?

[FullMetalAlchemist]

Does anyone know if ULE is put back in as the default scheduler? The 4BSD scheduler is stronger, and better in almost every aspect, but unfortanly not on the desktop where you'll notice the difference.
The reason I ask is that my less than experienced FreeBSD users that have cron'ed cvsup and buildworld might complain that performace when using VLC is decreasing; it's annoying to have to explain possible reasons without knowing.
The mentioned disabled PREEMPTION option seams to indicate this, but I'm not familiar with that option as I'm very happy with the old RELEASE which has another option (options SCHED_ULE).

Re:Anyone knows?

[drdink]

ULE has been set aside for 5.3-RELEASE. It is back to SCHED_4BSD for now.

Re:Anyone knows?

[sahtanax]

Sorry I can't answer your question on ULE, but thought I might point out that they now recommend against [freebsd.org] turning on PREEMPTION.

Re:Anyone knows?

[archen]

What kind of machine are we talking here? In my experience I really don't experience much of any slowdown what-so-ever using anything around a P3. It seems to me if you are using cron, then you should probably schedule cvsup when there is no activity. I find it sort of hard to believe the scheduler would make more of an impression if you set the jobs to 'nice 19'.

Re:Anyone knows?

[petabyte]

I'm using -BETA5 at the moment and well SCHED_4BSD fixed several issues that I was having with ULE and my laptop (it was almost unusable). On my desktop I didn't notice any change between them. PREEMPTION is disabled as of the last cvsup I did but I believe one of the reasons for going to 4BSD was so that PREEMPTION could be turned on again. Judging by some of the MFCs, it looks like they're tring to get PREEMPTION back for the GENERIC 5.3.

As for the the VLC thing, I've heard several people say that -BETA

Re:Anyone knows?

[FullMetalAlchemist]

It seams PREEMPTION is changes to how the VM works, i.e. it avoids a few NOPs and context switches.
The current implementation of PREEMPTION was designed for working with SCHED_ULE, not SCHED_4BSD, but as SCHED_ULE is causing trouble on it's own; and PREEMPTION is causing problems with SCHED_4BSD; I recomend only option SCHED_4BSD for servers.
At least try SCHED_ULE + PREEMPTION on the desktop, to see if you don't experience any problems.
The more people use SCHED_ULE + PREEMPTION the better, we want it fixed

Re:Anyone knows?

[DashEvil]

SCHED_ULE + PREEMPTION = Unusable for me.

Programs freeze and cannot be killed.

kill -9 does NOTHING to them, and they aren't zombied.

Oh, and then my computer arbitrarily decided to reboot.

Yea, it wasn't much fun. But the developers don't seem THAT interested in ULE + PREEMPTION crash reports because they are so frequent and easy to generate that the developers are having NO trouble getting test cases.

Honest Opinion?

[polyp2000]

Im using gentoo - running linux
kernel-2.6.8
(SMPx2way with low latency and pre-emptive enabled)

What are the main differences between the latest FreeBSD kernels and the linux ones.

A lot of people say FreeBSD is better because "its more stable" or "it has a more mature kernel" I've seen little evidence to substantiate these common claims. Whilst they may have been true some years ago are they really true today?

Apart from the cool things like the ports system and userland differences, licensing differences a

Re:Honest Opinion?

[FullMetalAlchemist]

Because it's a system and not a kernel?
To be honest, if you're happy with Gentoo, then use it; we don't really care either way, we use BSD because it's the best for us.

Try it, for 2 months at least, and then decide. If you like it, fine, if you don't fine.
I don't really care what other people uses, as long as it doesn't interfere with what I'm doing; hell, I've even started using Windows when .NET v1.1 was released, because *gasp* I like .NET.

Re:Honest Opinion?

[Brandybuck]

What are the main differences between the latest FreeBSD kernels and the linux ones.

It's not just the kernel that is FreeBSD's benefit, but the fact that the whole system is so well integrated. Even if you could, merely switching out one kernel for the better one would not provide you much benefit.

I haven't done any kernel work, and my kernel knowledge is limited to a Linux driver class and the new McKusick book. So take my comments with a grain of salt. But it seems to me that Linux tends to be very haphazard. There doesn't seem to be much evolution towards better subsystems, but instead replacement with alternative subsystems. For example witness the recent VM wars. There's also a distinct NIH (not invented here) attitude in Linux.

The FreeBSD kernel seems to be of a more "tried and true" design, with boring and unexciting implementations. The kernel improves by slow evolution. If NetBSD , OpenBSD or DragonflyBSD have done something right, FreeBSD is not shy about adopting it.

Pick up McKusick's "Design and Implementation of the FreeBSD Operating System" for more information(it's the successor to the venerable 44BSD book).

Re:Honest Opinion?

[0racle]

I wouldn't say that its a not invented here attitude, but more the fact the developers take a large amount of pride in their little corner of the kernel, and its hard to take pride in someone else's work when all you've done is implemented it. It goes toward the general feeling that Linux is more haphazardly put together, touting the individual parts as strong points whereas in the *BSD's its the finished product thats held high as a source of pride. You end up with a 'I did this myself,' compared to a 'we created this with the best ideas out there.'

Re:Honest Opinion?

[Brandybuck]

You pretty much said what I meant to say. If you have a cool new idea for implementing a subsystem what do you do? You're only choices in seeing it implemented are forking the project or starting a smear campaign against the old subsystem. DragonflyBSD did the former, and the myriad political battles in Linux attest to the popularity of the latter.

Re:Honest Opinion?

[Anonymous Coward]

You pretty much said what I meant to say. If you have a cool new idea for implementing a subsystem what do you do? You're only choices in seeing it implemented are forking the project or starting a smear campaign against the old subsystem. DragonflyBSD did the former, and the myriad political battles in Linux attest to the popularity of the latter.

Bwaahaha!! If you are trying to paint FreeBSD in a good light by bringing up DragonflyBSD, while also saying "Linux is too political", you were either born yes

Re:Honest Opinion?

[FullMetalAlchemist]

The Linux kernel development model works quite well.

There are however a few dark spots; the first come first serve attitude, like the old VM debacle for example; and all the other beta stuff that enters, like drivers for example.
The fact is that technical merits are always weighted by opinions, and opinions on lklm are strong.

These things will not bite you if you're sensible and don't update your kernels day in and day out; the few good Linux admins I know (maybe because I know few Linux users ;) are still on kernel version less or equal to 2.2.x.

It really doesn't matter what system you run; be it AmigaOS, Windows, VMS, BSD, Random Linux Distro, or OS/400, to name a few; if you know what you're doing you kan do pretty much anything.

Matt has and have always had a strong opinion on the weight on specific technical merits, namely simplicity.
You have to remember that Matt is originally from AmigaOS country, and his opinion therefor is strongly shaped by his love for its design.
While I agree with Matt, I also understand why the rest of the FreeBSD team picked another option; because it also, from their point of view, the best.
Matt's opinion is strong, and he refused to accept the choice of others, like he has done many times before; therefor we have DFBSD.
This is not bad or wrong, because we will have two excellent options to choose from, depending on how we weight technical merit with our opinions.

Re:Honest Opinion?

[drmerope]

The vaulted FreeBSD stability is mostly a matter of conservativism rather than the long history of the OS. Certainly much of FreeBSD was slowly evolved--and I think that applies to the 4.x series, but 5.x contains many aggressive changes to the Kernel. Thus, the history argument does not hold so much water.

Really, though, there are two kinds of stability. The uptime kind (really not that amazing in a UNIX OS, period.) and the system kind. FreeBSD is loved by administrators for its very well enforced POLA (policy of least astonishment). Point releases of the OS almost never break existing setups. E.g., 5.x will use use Bind9 and GCC 3.4 for it's entire existence. Should the originators of those programs neglect them, the FreeBSD project will step in and merge in fixes themselves. That sort of consistency is valued by many in critical applications.

A FreeBSD Kernel is not much different from modern Linux Kernels.

One of the big differences between the BSDs and Linux Distributions, though, is that the different BSD projects tend be distinguished by disagreements over system architecture--major design decisions at the Kernel level. Moreover, BSD projects tend to be focused on providing a complete, integrated system. Userland development and Kernel development always go hand-in-hand.

Re:Honest Opinion?

[YetAnotherName]

The vaulted FreeBSD stability is mostly a matter of conservativism rather than the long history of the OS.

Regardless of its source, it's that stability that made me choose FreeBSD as the platform for my firewall/NAT. Sure, I've got plenty of Linux, Mac OS X, and even a couple of Windows boxes on the inside, but that single FreeBSD system helps me sleep better.

(Oh, and I think you meant "vaunted", not "vaulted" :-)

Re:Honest Opinion?

[ArbitraryConstant]

"A lot of people say FreeBSD is better because "its more stable" or "it has a more mature kernel" I've seen little evidence to substantiate these common claims."

uptimes [netcraft.com]

You'll see a Solaris there. Occasionally. I don't think I've ever seen a Linux. It's exclusively BSD/OS and FreeBSD most of the time.

"Apart from the cool things like the ports system and userland differences, licensing differences aside- At the core level of the kernel what makes a new FreeBSD kernel better than a new linux kernel?"

Focusing on the kernel differences misses the point. The only stuff you'll notice is that Linux supports more filesystems, and FreeBSD has PF imported from OpenBSD.

The key advantage of FreeBSD is a very well tested base system. The ports give you a convenient way to add to the base system, and they tend to be quite well tested as well, but I've not seen the same level of quality on any Linux, least of all Gentoo (which basically doesn't do any regression testing and therefore breaks a lot).

Oh yeah. And the documentation. Linux docs are pretty bad. BSD man pages are famous for quality.

Re:Honest Opinion?

[molnarcs]

I had similar experiences with gentoo. And what really bothers me is that they advertised portage as having 'real' dependency checking in answer to a question of why not use ports. I was surprised to find out that (as of may this year) in fact gentoo didn't have proper reverse dependency lookup.

Also, dependency hell was changed to use-flag hell. When Midnight Commander installs (I don't know if it still does) XFree86 as a dependency, there is a problem. In FreeBSD, it is the task of the port maintainer to configure a port that is good for 99% of the users. Everything has sane defaults, while still preserving flexibility: most ports offer a nice ncurses based menu for switching off or on different knobs. In gentoo I had to know what a particular use flag does: in case of motif, I found the extremely redundant info: "This use flag will install motif on your system" - well, thank you very much! On FreeBSD I have to remember one thing: if I interested in configuring a port differently than the port maintainer did, I can look into the usually well commented Makefile (takes less than half a minute).

And don't start me on documentation. I saw a PR (problem report) about one description file of a port, coming from FDP (FreeBSD Documentation Project). The COMMENT in bold was considered a bug:

Description

The COMMENT for the x11/kdelibs3 port is (for KDE 3.2):

<b>This is the base set of libraries needed by KDE programs</b>

It is bad practice to start a COMMENT with 'This is the', since this could be
prefixed to almost all comments and has no informational value.

Fix

I suggest to change the COMMENT to:

Base set of libraries needed by KDE programs

Full PR is here [freebsd.org] - link might be slow because PR database is always overloaded. This is a paradigmatic example for the difference between FreeBSD and various linux distributions. I think this attention to little details and the general simplicity - read userfriendliness - of the commands, ranging from configuring your firewall to configuring start-up services is what makes FreeBSD so excellent. Forgot to mention: with ports, you can leave your puter building for the night. Even if some ports fail, building will go on, and at the and you will find a nice list of which packages succeded, which failed and why, if there were any. No skipfirst kinda stuff (which seems to a me workaround for the weakness of portage). Also make search key|name will give more info on basic level than portage would give on maximum verbosity. Makes looking for stuff easier.

So yeah, I know what you're talking about :)

ps.I studied literature (finished Univ. last year) - and switched to linux after a virus wiped out most of my very important docs in windows 2 years ago. I have never had any computer training. I loved RH 7.3 (my first linux), and I liked Mandrake even better (till 9.1). Then came debian, but still, there was this urge to try out another distro, and then another, and so on. Since I tried FreeBSD last year in september, I realized that this was exactly what I was looking for. Simple, user-friendly, very fast Unix-like OS, which is easier to learn (because of the documantation AND its consistency) for a newbie like myself than any other linux distro I have tried.

Re:Honest Opinion?

[ArbitraryConstant]

"Also, dependency hell was changed to use-flag hell. When Midnight Commander installs (I don't know if it still does) XFree86 as a dependency, there is a problem."

You think that's bad. Earlier this year Xinerama changed from a default to a use flag option. My WORKING CONFIGURATION spontaneously stopped working after a rebuild with no changes to the configuration, and I couldn't get an answer until I gave up on Gentoo entirely and then tried it again later this year. Then, when I was doing a fresh install I

Re:Honest Opinion?

[molnarcs]

lol. We are threading on dangerous grounds here, don't you agree? Two things prompted me to reply.

One: someone modded you troll, even though your criticism is specific and it is easy to look up your claims. Which leads me to my experience of the userbase. A ran into the absent nspluginscan problem, and searched the forums for an answer. That's how I found out that I need WITH_MOTIF. A user, just like me, found out this only after having compiled kdebase! I made sure that I didn't miss anything before compiling - and that's how I found the extremely useful description quoted above. Naturally, I was angry: there was no way to know before compiling a beast like kdebase that WITH_MOTIF is needed for plugin support. And of course no one can be expected to search for caveats on the internet before installing each port (or ebuild). To my astonishment, the user, who had to do a recompile just like myself, didn't share my exasperation. In fact, his reaction was: That's what I like about gentoo (referring to the use-flags, and ranting about the wonderful flexibility of the system!).

I only mention this because that somehow, modding your comment as troll reminded me of the absolute resistance to any kind of criticism (well, there are exceptions of course) on the part of the community. Saying anything against gentoo is dangerous indeed :)

Second: don't think (modders) that freebsd folks are antagonistic towards linux in general and gentoo in particular. Go to bsdforums, and search for the terms linux or gentoo, and you will find more threads that praise either of those than critical remarks. What really bothers me, however, is that I feel (both here on ./ and osnews) that when I am asked: why am I using FreeBSD, answering it is always like walking on eggs. Saying it is better in some respects (well, documentation is beginning to be accepted more or less) always results in attacks. So, for a time, I gave very careful and almost apologetic answers, full of remarks like this is just my personal preference, this is how I like it, OF COURSE THIS IS SUBJECTIVE, and so on. I am tired of this, so here is for burning some karma:

I LIKE FREEBSD BECAUSE IT IS BETTER IN EVERY ASPECT THAN ANY DISTRIBUTION I HAVE TRIED! Note the word: distribution. Not linux - linux distributions. It is as fast or faster than SLACKWARE, it has a kickass package management (pkg_add -r foo does exactly what apt-get install does - yeah, binary, precompiled, relatively up to date packages, and as a bonus, it has ports). It easier to maintain. It is straightforward. Documentation is unparallelled. Stable. It just works (out of the box, my usb mouse worked, without any ado and hotplugd or whatever). And most importantly: it is consistent.

One distribution might have excellent documentation. Another might be fast. Another might be simple (slack comes to mind). But I have yet to find one that has all of these. I use slackware occasionally, and except for lacking (or having to use 3rd party) package management, it comes close. But no separation of base and packages. Everything is dumped in /etc/. What does KDE do in /opt? Or gentoo: what does mplayer.conf or operarc do in /etc? Consistency is not the forte of gentoo.

ONE downside: linux distroes tend to have better hardware support (except for networking). Solution: I have fairly standard hardware, and next month, when I'm going to buy new components (a tv card) I would make sure that it works under FreeBSD.

Mod away!

Re:Honest Opinion?

[ArbitraryConstant]

"To my astonishment, the user, who had to do a recompile just like myself, didn't share my exasperation. In fact, his reaction was: That's what I like about gentoo (referring to the use-flags, and ranting about the wonderful flexibility of the system!)."

This is true zealots of other OSes (languages, etc) too. They think it's great because it suits them, and they can't put themselves in the place of another user that has different resources and needs. This is often true of people that are otherwise brilliant. Unfortunately, they get mod points sometimes.

I try to learn a lot of OSes and languages for this reason. I don't have a problem with something being different, or unsuited for my purposes. What I have a problem with is people that think their tool of choice is good at something it's not, or alternatively, that the task being discussed is not important.

Gentoo zealots think Gentoo is reliable and easy to maintain. I don't know how they can think that, but they do. I'd just shrug my shoulders and forget about it if they wouldn't lay on the evangelism so thick. I keep getting told things which I know to be untrue.

"I only mention this because that somehow, modding your comment as troll reminded me of the absolute resistance to any kind of criticism (well, there are exceptions of course) on the part of the community. Saying anything against gentoo is dangerous indeed :)"

Well, I did use the phrase "beneath pathetic". That's a bit pajoritive.

It's a bit irritating that the zealots of my OS (OpenBSD) of choice do the same thing. They seem to think it's a good desktop OS, which it's not. They tell people not to use Java rather than address the sorry state of Java on OpenBSD.

This is how I convince myself I'm not a zealot... I have a favorite OS which I use whenever possible, but I recognize that it's not a good desktop OS and use something else. I have a favorite language that I use whenever possible (Python), but some problems just don't want to be solved in Python, so I use other languages a lot.

Being a zealot will only cause you to miss opportunities. Knowing Python makes me a better programmer in C and Java. I can't stand Lisp, but it has the same effect. Knowing BSD makes using Linux easier. Expand your horizons people. Try something else.

I know 8 OSes (counting Linux once), and 12 programming languages. I can't stand most of them, but I'm a better programmer for it. And I found the stuff that was right for me. Python and OpenBSD weren't the first stops, of that you can be sure.

Re:Honest Opinion?

[molnarcs]

You are right about zealotry, of course. It just felt good for a second to say: [tongue in cheek]Hey my OS is better than your OS, muhaha[/tongue in cheek].

8 OS? whoops! I know 3: linux (more or less), freebsd, windows (98 - forgetting quickly). Recently I got accustomed with win2k (I have three OS on my puter: freebsd (use it 90%), win2k, slackware (5-5). I will try out XP soon (replace my win2k partition). I'm sorry to say I don't know any programming languages (few things in bash and even fewer in pyth

Re:Honest Opinion?

[ArbitraryConstant]

"firewall configuration"

Status of PF on FreeBSD [freebsd.org]

From what I can gather, it looks like FreeBSD is moving towards using PF as the default firewall (not sure about the timeframe), with IPF and IPFW retained for backwards compatability. It's available as a kernel module/port for NetBSD, and I believe DragonFly has imported it into the base system recently.

It can do some very spiffy things. It might be a good idea to try it out, particularly since you don't have to change OSes anymore. :)

Re:Honest Opinion?

[Homology]

It's a bit irritating that the zealots of my OS (OpenBSD) of choice do the same thing. They seem to think it's a good desktop OS, which it's not. They tell people not to use Java rather than address the sorry state of Java on OpenBSD.

Zealots are not particularly open for other "truths" than their own. This is not always bad, though. There are alot of zealots working for human rihgts. But I digress. The zealots you are writing about could just as well called narrow minded.

As for Java on OpenBSD : In g

Re:Honest Opinion?

[ArbitraryConstant]

"As for Java on OpenBSD : In general it's a license problem. The Sun Java License are very onerous. The OpenBSD developers are very dedicated about keeping their OS free, and in this I happen to agree with them."

Well, from the perspective of a non-zealot, it means I have to keep another OS around, which is non-optimal.

In reality, the number of things I need non-OpenBSD for is large enough that this is not a waste, but it's just an example.

Re:Honest Opinion?

[Homology]

Well, from the perspective of a non-zealot, it means I have to keep another OS around, which is non-optimal.

OpenBSD developers are really concerned about adding restrictions to existing licenses, and the effect this has on future free software. They really want to keep their OS free, and are making unpopular decisions based upon this. There are several recent examples, like the new license for XFree86 and the new Apache license. Both caused a fork.

Another example from last year is the new license [google.com] for

Re:Honest Opinion?

[ArbitraryConstant]

"OpenBSD developers are really concerned about adding restrictions to existing licenses, and the effect this has on future free software. They really want to keep their OS free, and are making unpopular decisions based upon this. There are several recent examples, like the new license for XFree86 and the new Apache license. Both caused a fork."

While what you say is true, it doesn't have anything to do with what I said. XFree86 and Apache are both in the base system. Java is not. There's plenty of things in

Re:Honest Opinion?

[Homology]

My guess is that if Sun Java had a more acceptable license, there would be more work on it. Some work are done on java, like native threads (for 1.3). I believe someone is working on 1.4_2

Re:Honest Opinion?

[ArbitraryConstant]

AFAIK they do charge for a source license, and it's probably a fair amount of work even though FreeBSD already has a 1.4.2 port.

However, as a user, it impacts me, weather not the reason is sound. I still have to keep a Linux box around despite the fact that they have a good justification.

PF alone is enough to keep me on OpenBSD, so I'm not getting rid of it, but it's irritating having to keep all these boxes around.

Re:Honest Opinion?

[Open $ource Advocate]

You'll see a Solaris there. Occasionally. I don't think I've ever seen a Linux. It's exclusively BSD/OS and FreeBSD most of the time.

The reason why is explained on the site, particularly the FAQ entry why don't some operating systems show an uptime above 497 days? [netcraft.com]. Another choice quote from the FAQ: "HP-UX, Linux, NetApp NetCache, Solaris and recent releases of FreeBSD cycle back to zero after 497 days, exactly as if the machine had been rebooted at that precise point. Thus it is not possible to see a HP

Re:Honest Opinion?

[Adam Avangelist]

I do not know how a FreeBSD BETA 4 release became into, "everyone point out the merits of FreeBSD over Linux systems, regardless if they are true" fest. It should be stated that netcrafts uptime benchmarks are unsubstantiated, there are many Linux systems that are up that are not on the Internet. Until 2.5 and above vanilla kernels Linux's uptime could not exceed 497 days; even still later 2.5 kernels according to Netcraft can not be judged correctly. Uptimes values can be faked, and benchmarks such as th

Re:Honest Opinion?

[JQuick]

A lot of people say FreeBSD is better because "its more stable" or "it has a more mature kernel" I've seen little evidence to substantiate these common claims. Whilst they may have been true some years ago are they really true today?

Apart from the cool things like the ports system and userland differences, licensing differences aside- At the core level of the kernel what makes a new FreeBSD kernel better than a new linux kernel?

There is a lot of hype on both sides of the fence here. Your skepticism is well founded. You asked several questions, which I will try to answer, but not without first adding some context. I think this additional context is needed because I suspect from your phrasing that perhaps you are asking the wrong questions.

You ask:

1. Is freeBSD more stable?
2. Is the kernal more mature?
3. What makes the FreeBSD kernel better then a new Linux Kernel?

The reason I am compelled to add some context is that your questions are lacking context. Value jusdgements like "better" and "more __ than" require context, in this case the context of utility. Better for what purpose, and towards which ends? More stable in the context of performing what tasks?

Another question that I have about the phrasing is stems from your mention of the ports system and userland as a preface to your question about which kernel is better. FreeBSD and Linux have different histories. Their user and developer communities have had different priorities. Each thus has different strengths and weaknesses in particular contexts. I would prefer Linux for certain specific uses and FreeBSD for others. The answer to your questions about stability and maturity will shed light on the rest.

Linux distributions emerged from a set of disparate components. BSD evolved from a complete whole. Starting from V6 and V7 Unix in the 70s, it has always been an indivisible whole, kernel, userland, documentation, the whole shebang. The technological strengths and weaknesses of each system, relate to this difference. The expectations and conventional practices of the systems users also are heavily influenced by these differences.

Linux users, developers, and administrators tend to applaud the rapid pace of development, greater number of new drivers and other features, and technical merits of certain specific aspects of the Linux kernel or userland. When hype is stripped away, I believe they are largely correct.

FreeBSD users, applaud, the stability, coherence, and ease of maintenance of their systems. Because it evolved from systems that were already in use, they always had an installed base, and many developers were motivated by keeping themselves and their existing users happy. The needs of system administrators were always well represented in the user and developer community.

Given a particular work load, and a particular set of applications, FreeBSD will outshine Linux on some tasks and be blown away by Linux on others. Each system is sufficiently mature that a well planned and well managed system of each flavor can run reliably for years. Yes, FreeBSD has historically been more stable and robust than Linux. It has a longer history, and its well managed release engineering process has kept the stable branches of the code remarkably robust and reliable over time. Given the rapid advancement of Linux however, the general stability and maturity of each system is quite high. If configured properly and deployed responsibly these gaps have narrowed and are largely moot.

The most compelling differences between FreeBSD and Linux are not terribly important if one deploys a single host and uses it directly. The difference is what is required to maintain the system over time.

FreeBSD is a unified whole, strongly influenced by the desire of administrators to keep users happy. As a result, the ongoing effort required to upgrade and maintain the systems is very smal

I'm sorry

[Nirbo]

I'd just like to apologize for the link being bad, I don't know how it happened... I copied it from my browser, and tested it, but somewhere along the lines, I must have not noticed somethign happening. :p, Thanks to the one who posted the link in a comment.

Re:ugh

[BasharTeg]

I agree that DragonFly has fantastic potential and that FreeBSD 5.x has been a long and bumpy road. However, when your post begins with:

Yeah, I know. It's a beta. However,

Stop typing.

Re: Stop typing.

[sp0rk173]

Here Here. I completely agree. When all of the underpinning of DFLY are implemented and they get their ports system up, I can see a lot of FreeBSD users moving to it in droves. Maybe then the FreeBSD team (well, what's left of it) will take notice.

Preemption still problematic

[molnarcs]

Someone mentioned here that perfomance gain, especially on the desktop, is noticeably with ULE. I agree. I've been using ULE since 5.1. The reason I switched back to 4BSD is that options PREEMPTION was said to be buggy with (and this was the main reason for making SHED_4BSD the default for this release).

However, when I tried recompiled my kernel with options PREEMPTION, I had two complete lockups under heavy load (compiling in the background, running KDE, etc). During beta4, I removed options PREEMPTION for this reason. Now I've put it back, and I experience similar problems. I got error messages repeatedly (and the OS stopped responding for a few seconds now and then) when downloading files (with roughly 500Kb/s from network shares. Perhaps the same errors would pop up in other circumstances of heavy disk I/O. The messages were like this:

Sep 20 21:37:59 mcsaba kernel: ad0: WARNING - WRITE_DMA no interrupt but good status
Sep 20 21:38:10 mcsaba kernel: ad0: TIMEOUT - WRITE_DMA retrying (2 retries left) LBA=123095148

After removing PREEMPTION, problems went away. Note that I used preemption with SHED_4BSD, not ULE!

On the other hand, 5.3 will be an excellent release. Other than problems with preemption (which I don't quite understand, I mean preemption, so I would be glad if someone explained to me what it does exactly!), the BETAS were quite stable. In BETA5, the old problem of floppy not working with ACPI on some chipsets (I have via) is solved. Start up time is very fast, I think it is faster than 5.2.1. (it starts up roughly 2x faster than slackware with 2.6.7 kernel on the same box.) Perfomance on the desktop is similar to previos 5.x releases (and I have a few problems with KDE 3.3 now).

Oh yes, another question. options PREEMPTION is listed under the SMP section of /usr/src/sys/conf/NOTES. Is preemption SMP specific? From the few things I've read (and the even fewer things I could understand :() I didn't think it was. Can someone explain this to me? And: does preemption help with latency? (I have problems with sound in some games, and I think they are latency related).

Re:Preemption still problematic

[Nirbo]

Amen. When my fd0 drive roared to life, it made my life :D. Prior to BETA5, I've never had it working with FreeBSD 5.x

The best sleeper feature

[skrowl]

Unlike many who have posted here, I actually read the article ;) Here's a great sleeper / not widely publicized feature: NDIS Binary Compatibility a.k.a. "Project Evil". FreeBSD i386 can use binary Ethernet and WLAN network drivers written to the Windows XP NDIS 5.1 specification. It is a little cumbersome to convert a NDIS driver into a FreeBSD Kernel Loadable Module (KLD)

Re:The best sleeper feature

[Billly Gates]

How stable is it?

I was under the impression it was just a hack and was alpha level at present.

Re:The best sleeper feature

[Moloch666]

Even though linux's ndiswrapper is older, I've found FreeBSD's works better. I have 3 comps, 2 Gentoo Linux desktops and 1 FreeBSD server at Beta5. All computers have the same DWL-G520 cards in them running with SuperG. All using the latest official DLink drivers.

The first desktop is just a video player running ndiswrapper 0.9. It works well except it does lose assoication randomly but seems get it back on it's own after maybe 1 minute max.

The second desktop is my main computer using ndiswrapper 0.10

5.3-BETA5 seems to work fine

[Anonymous Coward]

VS BETA3

Beta3 had problems with doing a make update and 5.2.1 on a multi-processor Xeon box was 'rebooting'. 5.0 and Windblows NT 4 on the same hardware lacked the reboot feature.

JDK1.4 even built under BETA5.

All the little fork'n processes look good.

Beta 6 is out

[whyne]

Beta 6 is out the mailing list post [freebsd.org]