Slashdot Mirror
FreeBSD 7.1 Released
Sol-Invictus writes "The FreeBSD Release Engineering Team is pleased to announce the availability of FreeBSD 7.1-RELEASE. This is the second release from the 7-STABLE branch which improves on the functionality of FreeBSD 7.0 and introduces some new features. Some of the highlights:
The ULE scheduler is now the default in GENERIC kernels for amd64 and i386 architectures. The ULE scheduler significantly improves performance on multicore systems for many workloads.
Support for using DTrace inside the kernel has been imported from OpenSolaris. DTrace is a comprehensive dynamic tracing framework.
A new and much-improved NFS Lock Manager (NLM) client.
Boot loader changes allow, among other things, booting from USB devices and booting from GPT-labeled devices.
KDE updated to 3.5.10, GNOME updated to 2.22.3.
DVD-sized media for the amd64 and i386 architectures."
frosty post
BSD be'z dead, yo! touch mah fro
I, fo' one, welcome our new black overlords. sho 'nuff!
A significant improvement on a crappy OS is still a crappy OS. No flame intended, but really... who uses FreeBSD anymore?
The next release of NinnleBSD is imminent. This will blow FreeBSD out of the water.
SUCK IT REPUGS
captcha: francs
Is there some sort of benchmark comparing FreeBSD 7.1 with other operating systems and distributions? I would be more than happy to run it on a couple of systems that I have hanging around but the user experience needs to be at least comparable to what I'm already running (kubuntu 8.10)
Slashdot, fix your code or at least hire someone who is competent at it to do it for you.
This is one of the better parts of this release. The lack of speed/clue on putting out both CD sized and DVD iso images has been highly frustrating, telling the users to basically "roll-their-own". I've already upgraded a few systems and things appear to be going well.
AMD is going to need all the help they can get to compete with Intel. I'm happy to see articles like this floating around, because it at least means AMD is getting some publicity.
Look at it this way. If Intel wins, then the consumer loses. Why? Because if AMD goes under that would leave only Intel left in the desktop/server market which means they won't need to invest more money to make better processors since they have no competitor to force them to and this obviously results in slower chips.
AMD better do good with there new Phenom II chips, or else I don't see them along for much longer. ATI alone won't keep them going for long.
"The difference between genius and stupidity is that genius has it's limits" - Albert Einstein
Kind of took them long enough...
FreeBSD kind of lost me with the 5 and 6 releases. I haven't tried 7, but maybe it's worth a shot again.
They would be wise to port WAPBL; it looks better than gjournal, seems to perform comparably to Softupdates (which are a data gamble), and doesn't have huge system requirements like ZFS.
Here are some descriptions of new features and links to benchmarks.
Or does it STILL kill the box every time it receives a fragmented packet?
"XML is like violence. If it doesn't solve your problem, use more." - Anonymous Coward
All the BSD's win for man pages that actually contain more information then "man pages are obsolete, please use the info documentation". In FreeBSD the entire core system has documentation. All of it written in the format god intended--roff.
Did you mention all the man pages are online and can be searched by version? Comes in handy when you are still using FreeBSD 6.1-RELEASE.
And did you mention the fact that BSD's aren't like Linux distros? FreeBSD isn't just a pooling of libraries and code from random people, the core of FreeBSD (shell and userland tools) are all done by the same large team. FreeBSD, OpenBSD and NetBSD are *cohesive systems*, not collections.
Want my year 2009 prediction? This will be the year of the BSD's in the data-center. There is a lot going for BSD based systems, and quite frankly the only reason I can see to go back to a random collection of tools and kernel code (i.e. a Linux distrubtion) is for running code that requires vendor support (Oracle, Dell, etc...). In 2009, I predict (hope) more of these big-name vendors officially support FreeBSD and friends.
As a lover of FreeBSD, I hope the guys in charge never try to "win the desktop". They'd never win and they'd stop paying attention to the stuff that makes it so good for servers. FreeBSD, and the other BSD's for that matter, belong in the data center. I'd argue the same for Linux, but that might get me slaughtered in these parts...
I've been using a USB-based FreeBSD5 image for a project for some time now. I wonder what they're talking about with USB boot support.
For every problem, there is at least one solution that is simple, neat, and wrong.
And don't be nervous about making contributions either. My first ports looked like shit, but the port guys were patient and over time I've gotten the hang of the system.
FreeBSD (and probably the other BSD's) are much easier to work on then the other guys. For starters, since you are using a *system* and not a collection of libraries, all your patches and bug-reports go to the same place. In other words, you aren't talking to "the website and the people who maintain the 'tar' utility", you are talking to "the freebsd guys". Your patch for "tar" goes to the same repository as the code for "libc".
Plus since it is licensed as BSD, you can actually contribute modifications and not worry about the nasty side effects found in other licenses. I've never contributed to a GPL project, but I've contributed tons to BSD projects.
Bottom line, FreeBSD is a great place to get your feet wet contributing to open source stuff. Good times.
But honestly, FreeBSD is a server OS. And for servers, it has pretty much any driver you need. Granted not all of it is vendor supported binaries (yet, but hopefully someday), but still, if you have a server from *big-co*, odds are good everything will work.
http://uptime.netcraft.com/up/today/top.avg.html
Not only are these the stats from perennial doomsayer Netcraft, but it's a double win since the top system is for Adult links. Enjoy your pics ...
BSD ain't even pining for the fjords yet either.
I've been sticking with the 6.x branch (6.4 most recently) as it's given me extremely reliable uptime with my Squid proxy servers. FreeBSD 7.0 excited me with their SMP updates and ULE scheduler aiding in performance, however I wasn't convinced that the long standing FreeBSD stability was there after reading a number of newsgroup discussions, and due to its immaturity. Now that 7.1 has been released, I'm going to start taking it more seriously for production use.
That being said, regarding some of the comments here, FreeBSD (in my opinion) is more suited to uptime, stability, and reliability in servers than it is to offering a performance oriented desktop experience. Want a good starter project? Try to make a FreeBSD stateful firewall with transparent proxy server (pf / squid) for your home using some spare parts you have kicking around.
Most of the serious service runners are blocking queries from such service as Netcraft. Getting on the top means that some kid will start probably DDOSing you for it. It's better to just remove you from that list.
Nobody in BSD land gives a shit who does what with code. That is one of the nicest features found in BSD systems--the ecosystem is pretty much free of open-source politics.
Nobody give a shit if you wrote your patch on a windows system and mailed it to the ports maintainers using outlook. Nobody cares if Apple, Tivo, or Cisco "locks up the code". In fact, better they do. The BSD licence makes it easy for those companies to contribute because they can use FreeBSD and contribute only the parts that aren't special-sauce. Companies *want* to merge their changes in with the mainline, it is expensive to apply patches to every version of FreeBSD. The BSD licence lets paid employees of these companies send in bug-fixes and patches without ensnaring the companies IP in a legal mess. Other licences have a tendancy to be all-or-nothing--either you hold on to your bug-fixes and merge them in for every version or you release your entire codebase to the world. BSD lets you pick and choose what bits can go into the world. Very flexible.
Bottom line... if Apple wants to use BSD code, who cares. Code is code. It isn't like it has feelings.
Does anyone know of the status on Java for FreeBSD. We mainly run Java enterprise applications on our servers, and although I like FreeBSD, we have been reluctant to using it due to "official" releases not existing of Java. There are "patchsets" for 1.6 and frankly, that word scares me a lot. We are currently going with CentOS for everything.
...devote my life to Open Source. FreeBSD in particular.
No real reason why FBSD. I just remember really liking Lehey's 'FreeBSD.'
Oh well, it's back to Visual C++ for me...
I am very small, utmostly microscopic.
The closest you'd get to an official java release (which I assume you mean is a binary compiled by Sun) is a binary package compiled by the FreeBSD guys. The only thing about FreeBSD packages is they usually lag behind the ports tree by several weeks.
Vendor support like what you are asking is one of the things FreeBSD and friends lack. My gut tells me that it won't be long before you'll see FreeBSD get enough mind share that companies like Sun start offering support.
So I'll revise my statement and say this will be the year that more big-name vendors officially support FreeBSD. By "vendor" I mean hardware guys like Dell, IBM or HP, not just software vendors.
I think while it isn't discussed much, GPLv3 made a lot of vendors think twice about Linux. My gut tells me that you'll quietly see more and more vendors back BSD based systems. There won't be much fanfare about it (the BSD world is pretty chill), but it will just slowly inch forward until most servers wind up running FreeBSD or OpenBSD instead of $RANDOM_COLLECTION_OF_CODE.
Just a hunch. Times are changing, and I could be wrong...
Thanks for the link of uptimes. It does my soul good to see nary a linux box in that list. I do have to wonder though is it because it is a bad operating system, or are their admins of a new generation that do not take pride in their work? Or is it because they lack the skill? Sorry, I can't work it out. If somebody has insight, please pass it on.
The biggest strength for linux is also its biggest weakness and that is there are a lot of developers. As a result, it lacks the polish of a commercial nix. And just so you linux zealots ask me to back it up, here it is. Take a look at rusty coat hanger abortion of files that is the network configuration files. It is not inutitive nor straightforward. It looks like some coders lost their shit on shroom trip and made it up as they descended into the abyss.
Here is a clue, use a file for your network configuration, maybe 3 or 4 tops, like Solaris.
The best slashdot sig around is
Linux is for people who hate windows, FreeBSD is for people who love Nix.
My current NAS is running FreeBSD 5.3, in constant use, and has: 10:31AM up 2331 days, 28 mins Kinda nice if you ask me, I also use it as a desktop environment on my laptop because it just "works" for me.
WTF Slashdot, why do I have to login 50 times to post?
I used to dual boot a BSD 6.0 system with linux but after it chewed up my ext2 /home directory on a couple of occasions I just stopped using it. Not worth the hassle of restoring from backup just to use an OS that offered little over Linux aside from quicker bootup/shutdown times.
There are more important things in the world then how well an operating system does in some assholes random benchmark. If you are standardizing your servers around an operating system based solely on "speed", I question your abilities as a server dude.
I'll just name one thing, out of many, that are vastly more important than "speed". Stability. No, not "never blue-screens". I'm "does the maintainers of the system make major changes in every single release and then stop supporting older releases". Under this definition of stable, FreeBSD wins over linux hands down. Especially after the "we can't be bothered to maintain a stable branch of the linux kernel, so we will add new shit in with the old all the time". You might get a dozen exciting new bugs and security fixes when you "upgrade" between 2.6.1114492 and 2.6.1114493. In fact, this was one of the major reasons for me dumping linux in the first place. The 2.4.x kernels are the last stable linux kernels out there.
That is just one example of something more important than "passes 4*10^30 fps in WoW" benchmark.
As for security? Which is easier to audit and verify? A random pool of code and libraries distributed across hundreds of websites and maintainers, or a cohesive operating system whos entire codebase is in exactly one place?
been a freebsd user since 4.x days.
I use bsd to run my mail, antispam, dns and other public web services.
I'd LIKE to also have it be a fast samba server but for some reason, samba on bsd really SUCKS. why is that??
my similar hardware linux box runs circles all over bsd on samba. that's the last hold-out, really, in wanting to go all-bsd at home.
is there EVER going to be equiv speed on freebsd as linux has, for smb?
--
"It is now safe to switch off your computer."
Woo! Finally got to see what everyone else has been talking about. Not signed in currently at all. I'm not a paid subscriber anyways.
And there's about 50 posts already.
As someone who both enjoyed discovering GNU Info (as it was about the only part of the GNU platform I could run on a 2MB Amiga 1200), and also enjoyed discovering the quality of FreeBSD's man pages, let me give another perspective:
There's absolutely no reason not to use HTML for documentation these days. There are plenty of lightweight text-mode browsers that would suffice in emergencies or during ssh sessions, but also nice desktop apps that would let new users browse them and feel at home. More importantly, it supports modern features, like links to the actual organisations online who support a particular app, or where bugs can be reported, links to email, diagrams, unicode for multilingual support, screenreader support, etc.
Yes, manpages can be nice, and coherent, quality documentation is important. GNU's horrible info browser is certainly not up to it. BUT... let's get with the times. There's no point advocating man pages in the modern world. If you want good docs, argue for good docs in modern formats, not old formats that happen to sometimes have instances of good docs.
Benchmarks between competing free software projects? Don't be silly! Next thing, you'll be advocating some sort of sane system, like choosing the best of breed technology based stats like benchmarks, and uniting behind it! Think what kind of chaos Free Software would be in, if everyone decided that OpenGL was THE low-level graphics layer, that gstreamer was THE codec API, that Vala was THE high-level language, that Git was THE modern version control system, or that FUSE was THE place to develop filesystem stuff. Why, you'd have a straightforward stack, with very little bloat, and tons of people honing a single implementation.
Pandemonium, I tell you.
As in going into a port and "make -> make install"?
If you are go grab "portupgrade" (/usr/ports/port-utils/portupgrade, I think). Portupgrade will do the "make" crap for you and has the side-effect of doing a "make clean" when it is done. It has some other nice parts like letting you set all the config variables in one file as well has helping you do crazy gentoo-like dependency swaps.
PPS: "make portsnap" while you are at it and then put it on a cronjob. cvsup is for people who are gonna fuck with the ports tree.
There is a difference between "You guys aren't playing fair..." and "our operating system is your religion, either embrace it or go away".
If somebody like $VENDOR_X takes and takes but never contributes even minor shit like bug-fixes to kernel code, they should be called out. But unlike other, more political organizations, you will never see an Anti-$VENDOR_X clause added to a BSD license. That is the important bit.
BTW, one big peeve in BSD land is when the GPL guys will take BSD code like drivers. The GPL license will "infect" any modifications and prevent those changes from being send back to the original BSD code. Kind of a tease, don't you think?
2009 is the year of the flow chart, which will then bolster the popularity of FreeBSD by definition.
df -h
(and boy I'm posting in this thread ;-)
For those who've never used a BSD system but have used Linux, be prepared for the command line to work a little different. BSD utilities are often way more picky about the ordering of arguments.
With the GNU tools, "chmod 775 * -R" will recurse down a tree and set everything to 775. "chmod -R 775 *" will do the same thing.
In FreeBSD, only "chmod -R 775 *" will work right.
In BSD userland, the patten is almost always command [arguments] [strings of goo]. In GNU land, you can usually interchange [arguments] and [string of goo] and get the same result. Some will argue that only the BSD way is proper and the GNU way is sloppy. Whatever your feelings are, if you've gotten used to being sloppy about ordering, it will take some adjustment to get used to BSD tools.
The good news is the "proper" way will work on either set of tools.
...And it's been a night-and-day difference between administering Ubuntu and FreeBSD boxes.
The management tools and fantastic attention to integration have been a true pleasure (example: install portaudit and you'll automatically get security alerts for your installed ports daily). The documentation is excellent; see "FreeBSD Handbook" for several hundred pages of great material. Everything I've touched so far just feels so... COHERENT.
I'm basically done with Linux.
Possibly. There was a bug in directory listing for large directories where entries were deleted dating back to something like 4BSD and inherited by all of the BSDs. It was found and fixed about six months ago - it turned out the Samba guys knew about it and had a work-around in place, but the work-around made directory listings very slow.
I am TheRaven on Soylent News
Does it have Xen virtualization support? It would be nice to run it as a guest OS for testing on a server without using emulation.
- Michael T. Babcock (Yes, I blog)
Everything in the ports tree is essentially random crap. The only thing FreeBSD does is wrap the source code with a (really nice) build system. Ports aren't "stable" the same way the core is. That said, a lot of the big-name stuff like apache has separate ports. For example apache-1.3, apache-2 and apache-2.2 have separate ports (I think there is a port that follows the trunk too).
The difference between the BSD's and Linux's are in scope. In FreeBSD there is a whole lot more junk that is maintained by a single group then in most linuxes.
But still, you are correct in the "real applications" are all ports.
I think it is fixed - give PC-BSD a try on your at home box to test it out. - pcbsd.org
And this new FreeBSD release can run on Dolby Surround 7.1 systems?.
Wow. Do you have a link to a description of that bug? I'm curious.
Please correct me if I got my facts wrong.
How many passed BSDs you can see on http://www.opengroup.org/openbrand/register/xy.htm ?
The certifications go per version. No *BSD is certified Unix at this moment, at least not for Unix03 or Unix98, which matter.
It was covered on Slashdot. Here is a detailed write-up. Basically, the problem occurred when seeking to the second entry in a block when the first one had been deleted due to some mismatch between what the kernel did and what libc did. The bug was around 25 years old, and was fixed in May. It only occurred in a very small number of cases, but these cases were common enough for the Samba team to have encountered them and worked around them.
I am TheRaven on Soylent News
Having had the pleasure of using various versions of Solaris & HP-UX I have to laugh at your naivety and faith in "Commercial UNIX"
Yeah, we could call them things like /etc/sysconfig/network-scripts/ifcfg-eth0, /etc/resolv.conf and /etc/hosts. Oh those silly Linux developers!
My point was that things are different. Personally I actually like the more sloppy approach used by the GNU utilities, even if the programmer in me disagrees.
Another thing that takes a little getting used to is the device names are different. Rather then "hdd0" and "eth1", they are named after their kernel module - maybe "ad0" and "ee0". I'm sure there is a good argument for either approach, but it is something that takes adjustment to.
I've got a couple problems:
1) How do you deal with an HTML document that links to something outside the filesystem? As in, what if I'm offline?
2) You can do a hell of a lot in HTML that you cannot render on a console. What if the document you wrote uses an image or some javascript?
3) You need to make it compatible with "man". While 'info' sucked big time, one of the suckiest parts was that it fragmented the linux documentation. Whatever you propose must exist when I type "man joeblow". Failure to do so will repeat the mistake made by the 'info' dudes.
Problems 1 & 2 need to be addressed because one of the use cases for documentation is when shit hits the fan and your net doesn't work and your only access is on the console via 9600 baud RS232.
Problem 2 may not seem that important for some things (developer docs, I'd say), but for other things like shell utilities the documentation must be presentable in the same medium as the utility. I have no problem with web-based programmer docs, but I do have a problem if some jerk wants to use an image in his documentation for 'tar'.
There is another hidden issue. HTML isn't a standard anyway. Which HTML version? How strict is "strict"?
It took me over six months to figure out how to install 4.10 seven years ago, but you taught me a lot, and here I am today - nearly indispensable to my employer - because I'm the guy who "figures things out" (albeit much quicker now).
What?
But you have to admit it is kinda a tease. Here you go and modify my code and then dangle your changes in front of me and yet I can't use them. At least when you put it in your Fortune 500 breath-mint testing software I can't see your changes--out of sight, out of mind. With the GPL stuff, I can see but I can't touch :-)
Bottom line is no, I dont give a crap if you never give me a line of code in return. But I still am human and am thus subject to fuzzy, non logical things like culture and being nice to your fellow man.
But not a whine. You are free to be pedantic and interpret it as one though... others seem to relish in the opportunity.
A free, open source operating system with a huge amount of the code maintained by a single company which can provide official, professional support.
Sun has come out with two of the big things BSD is trying to import: DTrace and ZFS. I predict Crossbow will be next.
The drivers aren't quite there, nor is the massive package repository. But assuming Sun will not abandon its OS dept. or get sold off and thus fold the project, I believe Opensolaris has a bright future.
That said, as a person who has gotten into sysadmin only in the past two years, the OS that got me interested in Unix was FreeBSD 6. I still like its filesystem organization design better than the Linuxes, as well as its reputation for stability.
If only Ubuntu had based itself of FreeBSD instead of Linux... Sigh. It could be so much more powerful, and unrestricted by RMS to what software and code to use to improve itself.
What other choice did they have? All the other unix variants are proprietary.
The best way to check for fixes is to run 8.0/current. I'm using it on my laptop fulltime, and hasn't given me any problems. Won't recommend current on servers.
Good god, he was asked a question, and he answered it. It wasn't even a question about code, it was a question about funding... and he has a real point. The only way I was able to get one company I worked at to contribute ANYTHING to OpenBSD or FreeBSD was by writing up purchase requests for media and shepherding them through channels.
Also, that quote does not carry Theo's tone of voice. From my experience with him, I would guess that he was laughing all the way through that last sentence... it comes out as a rant in print but that's not how he sounds in person.
FreeBSD is the only distribution, other than Solaris, to have ported and implemented the ZFS filesystem (and no, a FUSE port doesn't count).
I've been looking forward to build a file server for personal use, and I'm eager to try out ZFS, which really puts FreeBSD high on my small list of candidates for an operating system. I'm going for consumer-grade hardware, and I'll be experimenting with stuff like using CompactFlash cards to store the OS.
OpenSolaris was my initial choice due to its higher maturity on the ZFS implementation, but I feel it's too constraining. I tried searching around for information about installing the system on flash mediums, information about wear-levelling, filesystems for flash media, and their forums and mailing lists fall short on these topics. The OpenSolaris installer doesn't even allow one to customize the installation, forcing me to install X.org, Gnome, and a ton of other stuff. No thank you, I'd very much like my file server to be command-line only, and to be smaller that your 3.1 gigabyte minimum for an installation.
As soon as I feel that FreeBSD's implementation of ZFS is stable and feature-rich enough for my needs, I'll definitely be rolling a file server with it. And I don't care if Netcraft disagrees with my decision; I really do feel BSDs deserve more and more notoriety these days.
Many who use GPL do so because they dislike proprietary software. Yet when it comes to integrating with other open source software, GPL'd code amounts to what is essentially proprietary software. Might as well ship a binary blob, it is just as useful.
Uhmmm, Postel's Law, users (transparently) love it, programmers (selfishly) don't ;)
i found that my poor samba speeds came from the nic on the windows machine. once i'd swapped that out for a spare intel 10/100 pro i had kicking around i was getting the expected 12Mbyte/s.
Music is everybody's possession.
It's only publishers who think that people own it.
Fuck Beta
~John Lenno
Vitriolic comments - check
GPL/BSD flamewar - check
Bashing Linux - check
Bashing RMS - check
Adoration of ZFS - check
Hardware support discussion - check
I think we have all bases covered by now.
Is BSD becoming the new Apple?
Those who would give up liberty to obtain working drivers, deserve neither liberty nor working drivers.
I wish Vmware or Sun (virtual box) would start supporting freebsd 7 for client-side tools. It would make running and playing around with it a lot easier and fun.
I find it really annoying how BSD crowd is often stating that BSD comes from single repository (ergo something good) as opposed to Linux where libs come from whole variety of places (ergo something bad), repos and developers. This is simply not true.
First of all, comparing FreeBSD repository to all linux distros is false. Compare FreeBSD to _single_ distro, whichever distro. Then you'll see that the code is coming from single repository too. OpenSuSE maintains its own patches (and therefore repo) for software. Ubuntu does the same. Pretty much every distro that patches original software does maintain its own repo. Just like FreeBSD. Because, FreeBSD's repos is really not OpenBSD's right? Or NetBSD.
So if you want to compare FreeBSD to Linux, compare it to single (any) distro.
Secondly, Linux distros are not a random collection of libs as some FreeBSD fanbois would put it. Pretty much all distros use SAME software, some with more or less custom patching, mostly branding (like OpenSuse for instance). They are doing it for various reasons, but the core functionality of each package remains the same. You can use Open Office equally the same, with same result on any distro. So it is far from being a random collection of kernel + software.
It is good to hear some of the Linux distros are heading in that kind of direction. It has been a while since I've looked at the state of Linux distros (besides Ubuntu).
Quite frankly, once I made the plunge into FreeBSD, I've never really looked back. I'm a happy camper with FreeBSD on the servers and Windows on my desktop. I'll spare you which version of Windows I'm using :-)
I've been deploying FreeBSD in the last 10+ years, mostly as chief architect/CTO in startups, and here are some of the reasons I still opt for it today - currently about a dozen servers in cages serving cumulatively 200 Mbit/sec, all non-http proprietary protocols and software (we have nothing web-related,) software-based redundancy, MySQL, etc:
1. Stability.
In the last 10 years I had kernel panicked once, due to a driver problem.
This needs some discipline and patience:
- We introduce a new RELEASE only after we have been running it for at least 6 months on our test systems, and then only when it becomes cost-effective over patching the previous one.
- Everything we run is compiled from the source, to the last executable, on the target architecture (we have several.)
- We create our own automated installation media, certified for our targets, which results in half-page instructions and few CDs that a retard can install in less than an hour - from bare metal to the running server.
- We never run any GUI on servers. All access is via SSH (which is actually the only open port apart from our proprietary protocol ports) and engineers use a wild range of development/interface machines. Some even use Linux.
2. Maintenance
Usually, for each release, there are few problems that we have to solve.
Most of the time it's patching network drivers (as hardware is always ahead of the release). Search for the problem, figuring out what to do, testing the patch, incorporating it into our installation process - all of that never took more than 2-3 days. At this point I am willing to bet that 7.1 will run on all our servers with less than a week of extra work (currently we are on 7.0. 7.1 will get phased in in August or later.)
Outside kernel, there are sometimes application bugs, generally nastier than kernel problems. MySQL had one with faulty locking on multiprocessor machines, which we patched in May last year (it took about a week), and which only recently found its way to MySQL release.
All taken into account, it's usually less than 2 weeks of work and about 6 months of test run for each new release.
3. Performance
We never ran into a performance issues, simply because we add a server when average CPU/io usage goes over 20%. Our job mix appears to uniformly load CPU, network interface and disks. A server costs $4-8K and it's a non-issue for the kind of business we run. So maybe we are just lucky here. The only real issue is bitmap rendering that we do in batch once and a while (can take a week on *fast* 4-core server), but we do this once every few months.
None of us, except one strange guy, uses FreeBSD as workstation machine. It sucks to force Linux GUIs onto FreeBSD (even the weirdo runs some simple screen splitter), so I have no clue about that aspect.
4. Elite status
They used to say, "Would you rather date crowd in PC or Mac section in Frys?". Well, you get the idea.
And I find it interesting how companies can "exploit" GPL the way you describe. Especially when you get contributors to assign their copyright over to the maintainers. There is no way to "exploit" a BSD license that way.
I'd actually be curious to know what the official stance is on your use of the GPL. While I personally dont give two shits, I wonder if the FSF guys have a opinion.
its not my windows system.
on my production (and even test) systems I use intel epro1000 chips, even integrated on the mobo on pci-e if at all possible.
my xp desktop to linux: very fast
my xp desktop to bsd: lots of pauses and wait states (you can see it on the switch leds - being dark MUCH of the time)
linux to bsd: also slow
everything to everything via nfs: lightening quick (using MS's unix 'stuff' on xp). all switches are gig-e, nothing using jumbos.
smb on freebsd is just NOT very fast. I seriously dislike having to use it and am thankful that at least the MS software download for xp enables NFS and I can mostly ignore samba now. but - its STILL a flaw that bsd, out of the box, can't even come close to linux (ootb) on samba.
--
"It is now safe to switch off your computer."
I don't think this was the cause for the slowness I see with bsd/smb. this does not sound like a common thing that would be affecting my traffic.
one test case I use: on my xp box I'll mount a smb server's disk and do a cd or dvd rip across the network.
when I rip to linux, the network led is solid-on and the file completes in minutes. if my remote server is bsd, though, the dvd rip might take half an hour or so. HUGE diff.
I doubt its this dir bug that I'm hitting. I think the perf. issue is something more fundamental. it might be bsd's fault or smb's fault - but the combo has NEVER been 'high performance' to me, sad to say.
--
"It is now safe to switch off your computer."
Where is the Perl 5.10 port!!!
FreeBSD is for people who love Nix
No, FreeBSD seems to be for people who hate Linux. FreeBSd fanatics are consumed by their hatred of Linux. It's understandable, I suppose, since Linux basically stole FreeBSD's thunder
I am a die hard Linux fan and I agree with what Linus Torvalds says about BSD. All BSD guys, users and system developers are nothing other than a bunch of Masterbading Monkeys.
I call bullshit.
On a 100Mb network, the expected maximum speed is indeed 12MB/s. However, this does not take into account the overhead for the protocol stack, this is raw Ethernet speed. With TCP/IP overhead, you'd drop down to about 10MB/s, and then you get into the SMB overhead, which is worse. No way you're getting SMB file transfers at 12MB/s on a 100Mb/s network. It's simply physically impossible.
Mart
"I know I will be modded down for this": where's the option '-1, Asking for it'?
FreeBSD 7.1 still ships ports/lang/neko while the software never worked on FreeBSD (nor on any other *BSD). And yes, there's an open bug for ages.
Seriously, what keeps me off FreeBSD is the fact that it ships things that weren't tested. Remember when a commit to -STABLE broke Realtek drivers ? Remember when the ports tree had a completly broken pecl-APC, a critical component used by many web sites (and it lasted a long time before being fixed, waiting for an upstream update) ? That's not old, that with FreeBSD 7.0. Oh and of course the broken PHP 5.2.7 also got commited to ports.
Can't FreeBSD folks test things instead of rushing and remove unmaintained things or things that never worked and that they aren't able to fix?
{{.sig}}
I updated yesterday. Pulled down the sources and built world and kernel. All during this my system remained usable. I didn't have to upgrade any ports. After build I dropped into single user mode, installed, then rebooted. When I was back up, I noticed no differences... and that is a *good* thing! All operating systems should be like this (although with varying degrees of hands-on-ness).
Don't blame me, I didn't vote for either of them!
I'm particularly happy about 7.1, as it has introduced support for booting from GPT partitioned disks. At long last, the tired old MBR and BSD label system can be thrown on the trash-heap of history. Since we have mostly macs with one FreeBSD server in the house, the shift to GPT was mostly already done. Now the only media not GPT partitioned in our house is read-only media or small flash drives (SD or thumb).
Some advice for anyone considering it:
1. Don't use the 'gpt migrate' command. I did my transition by buying a new hard disk, but once it was complete and I was happy with the results, I experimented with the old drive. The migration went poorly. For one thing, you need to have enough space unallocated at the end of the disk to fit the GPT protective copy, but even then, it didn't migrate my setup properly.
2. You need to use 'gpt boot' to make a boot partition that contains the first stage loader. This also writes boot code into the MBR that loads this boot partition. This loader, in turn, will load the actual loader from the first UFS partition and from there you're off to the races. At some point, you may transition to an EFI based machine and will need to replace this boot partition with an EFI system partition, into which you will add (to be written) an EFI loader for FreeBSD. Changing this over will be made much, much easier if your boot partition and swap partition are next to each other. Transitioning will be a simple matter of deleting the swap and boot partitions, creating the EFI system partition, then creating a new swap from the remaining space.