Some people already submitted this news last week, but it wasn’t until today that it became official: the FreeBSD team has announced the release of FreeBSD version 9.0. As you may expect from the major version number change, this is releas eis packed with new stuff.
I’ll make no secret of it: the BSDs really aren’t my area of expertise. I’ve always seen them as the more ideologically developed alternative to Linux, which tends to take a more practical approach to development. Considering my lack of knowledge, here’s what the team itself considers the most important new features and changes in FreeBSD 9.0.
- A new installer, bsdinstall(8) has been added and is the installer used by the ISO images provided as part of this release
- The Fast Filesystem now supports softupdates journaling
- ZFS updated to version 28
- Updated ATA/SATA drivers support AHCI, moved into updated CAM framework
- Highly Available Storage (HAST) framework
- Kernel support for Capsicum Capability Mode, an experimental set of features for sandboxing support
- User-level DTrace
- The TCP/IP stack now supports pluggable congestion control framework and five congestion control algorithm implementations available
- NFS subsystem updated, new implementation supports NFSv4 in addition to NFSv3 and NFSv2
- High Performance SSH (HPN-SSH)
- Flattened device tree (FDT), simplifying FreeBSD configuration for embedded platforms
- The powerpc architecture now supports Sony Playstation 3
- The LLVM compiler infrastructure and clang have been imported
- Gnome version 2.32.1, KDE version 4.7.3
As always, architecture support is outstanding, so no matter if you have an amd64, i386, ia64, powerpc, powerpc64, or sparc64 box, FreeBSD 9.0 will happily run. One note, though – a problem was found with the DVD images for amd64 and i386; new images have since been uploaded, but if you downloaded one of these images before this announcement, it’s wise to re-download the image, or verify its checksum.
Have fun, and FreeBSD users, let us know how it goes!
BSDs hate the GPL. One big goal for FreeBSD 10 is to be GPL free (in base) http://wiki.freebsd.org/GPLinBase
I think that is a last resort effort for gaining any traction against linux, which won’t work, but it is still nice to have a fully BSD(like) licensed base. Took them long enough.
More info:
http://wiki.freebsd.org/GPLinBase
No, it’s not a “last resort effort to catch Linux”. It’s a pragmatic response to the GPLv3, which has prevented a lot of GPL’d software from being upgraded in the FreeBSD source tree.
The GPLv3 is the reason that FreeBSD is stuck with GCC 4.2 (the last GPLv2 release). Same with many other GPL’d apps in the source tree — they can’t be upgraded since the projects have migrated to GPLv3.
Hence, the big push recently to get non-GPL’d versions of these apps developed and imported.
Has absolutely nothing to do with “catching up to Linux”.
Edited 2012-01-12 23:36 UTC
I don’t think this has anything to do with gaining traction against Linux. Linux is and has been more popular than *BSD for years, and I think the BSD devs know and accept this. In fact, not having to compete with linux gives them the space and time to focus on pursuing goals of a more philosophical nature, like replacing all BSD incompatible licensed software in the base install.
edit: fixed typos
Edited 2012-01-12 23:39 UTC
You need to watch the LLVM video regarding FreeBSD – it has nothing to do with ‘loving’ or ‘hating’ anything but the reality that there are some companies that use FreeBSD which have a no GPL3 or GPL policy and thus it is one area that needs to be addressed.
http://llvm.org/devmtg/2011-11/
Great some strange quicktime movie that my mplayer won’t play (Apple.com videos work)
So I guess it is true, every FreeBSD developer uses OSX as their desktop.-
Works perfectly fine in VLC – MOV is a QuickTime container where the source code is open source and freely to be used by anyone, it can hold pretty much any sort of CODEC inside it with the most popular being h264. There is a bug in ffmpeg/mplayer that has existed for years but there has never been the motivation to actually fix it which is why you’re having the problems you’re having now. In an ideal world the LLVM maintainers should have stuck with m4v/mp4.
No idea what bug you’re referring to. The movie in question works fine with FFmpeg and MPlayer.
Your ideal world comment makes no sense. MP4/M4V and MOV are both QuickTime.
It plays fine here with VLC, Xine, MPlayer and ffplay.
As if MPlayer is the only media player. Get a clue.
Same here. Unfortunately, Linux still sucks at playing non-flash video. It’s very hit and miss.
Your comment makes no sense and it’s also backwards. Last time I checked the kernel isn’t a media player.
Can you play the linked video? If so, what obscure codec did you install to allow it?
Regarding your silly distinction between kernel and OS, that’s a classic FSF double-talk ploy. Linux distros have never been very good at dealing with the variety of codecs out there. And much of it is not that they can’t do it, but that many of them won’t. The more aligned they are with FSF the less likely it is that they will be able to play any particular video without a lot of fiddling by the end user. Denying that obvious fact doesn’t help the situation.
I will ask you again, for emphasis. Can you play the video? And if so, what did you have to do to play it?
I’ve tried playing it and it works with MPlayer, VLC and ffplay, it also works with any other player and distro I try.
Your problem is called PEBKAC.
Edited 2012-01-13 04:26 UTC
It doesn’t work in Totem or VLC here. Both are dependent upon the specific codecs and libraries installed. You can hide your head in the sand all you want. But few users are willing (or capable) of jumping through the hoops you and I might be willing to jump through to watch a video.
Getting Windows for free, or even paying for a Mac is preferable to most. You know that as well as I do, as much as you might want to deny it. It’s an area where we need to be doing a lot better than we are.
Uh, I navigated to the video in chromium it opened mplayer in browser told me it need to install some codecs, then found the codecs and installed after I authenticated and approved. It could not get any easier in Fedora 16.
Step 1: Install a popular linux distro
Step 2: Watch almost any video ever made
Whoo! All those hoops. I can’t believe I was able to wrap my mind around it.
No shit, non-flash video in linux is easier than it is in Windows for me (until I discovered VLC at any rate). And even where flash is concerned, use a distro made in the last half-decade and you’re golden.
Well the blame is being put on FFmpeg but it is actually with the Operating System in question and the idiotic defaults being utilized. But it definitely does not affect every Operating System utilizing the Linux kernel.
*sigh* Look. If we’re doing everything right… why has Linux desktop use been stuck at ~0.75% market share for so many years despite being free, being Free, and having had lots of press and lots of enthusiastic public advocates?
http://gs.statcounter.com/#os-ww-yearly-2008-2012
This graph only goes back a few years. But it’s quite in alignment with what I recall from the late 90s, early 2000s. The number I recall from ~2000 is ~0.5%. So maybe that’s a tiny bit of progress over a 12 year period.
But for all practical purposes, desktop Linux is going nowhere fast. Hiding our heads in the sand and complacently saying “Everything is fine. Works for me.” is obviously not an effective strategy.
Hurl insults and spit venom at me all you want. I don’t care, particularly. But it doesn’t change the cold, hard facts. If my efforts get a few more people to see the reality and ponder a solution, I’m happy.
-Steve
Edited 2012-01-13 19:02 UTC
Both the .mov and .mp4 played perfectly in VLC 1.1.13 and MPlayer on my Arch Linux system.
That said it’s about the codecs, afaik VLC relies on ffmpeg and as such it depends on what codecs your distribution enables by default in ffmpeg which in turn depends on how they percieve the threath of patents. Arch linux has a pretty much anything goes attitude and generally enables patented tech (codecs, 3d) but it’s also a volunteer driven distro and likely not a target for patent infringement suits.
Still every distro I’ve come across has the option available for the user to explicitly enable patented tech (often all you have to do is enable a repo) so it’s not a big problem.
Yes I can and I didn’t install anything. It just worked. The video is using H.264.
It’s not silly. It is reality. Linux is a kernel. Red Hat, Debian, SuSE, etc. are Operating Systems.
The project needs an up to date and maintainable compiler/toolchain that meets the projects requirements and with GCC past 4.2.1/binutils past 2.17 that is no longer possible. It’s a much bigger issue than just some companies having an issue with GPL’d code.
Stop with the bs, the ONLY reason the BSD’s didn’t upgrade past 4.2.1 is because of the GPLv3 licence and as others have already stated this was because of companies supporting BSD having a ‘no GPLv3 policy’ (most likely due to the TIVO-ization clause).
Also this relates to what the BSD’s ship with, all later GCC versions have been readily available through ports so those who have no problem with GPLv3 and wants to enjoy benefits of later versions like better optimization/faster compilation can get them.
It’s not bs. GPLv3 is not acceptable for the base OS. I can’t help it if you cannot accept that the project has requirements and actually sticks to the requirements. It also goes beyond just the license when
it comes to using and maintaining the compiler.
The base OS compiler is what counts the most. It’s a concept that seems to be lost on Linux people.
Yes, GPLv3 was the problem.
What problem occured with using GCC after 4.2.1 that was NOT GPLv3, please elaborate.
Again, what problems beyond 4.2.1 did GCC introduce apart from a licence shift to GPLv3?
Answer this question for me.. are you a *BSD developer never mind one that actually works with the compiler/toolchain and helps to maintain it?
Edited 2012-01-13 03:47 UTC
Are you?
You are the one who claimed that it was not due to GPLv3 but compiler toolchain maintenance problems past 4.2.1 which prevented use of later GCC versions and yet you haven’t been able to point out any such problem, and also given the fact that later GCC versions are available through ports and from what I understand are capable of building the BSD’s then your statement holds no water.
Unless of course you can point me to anything which backs you up?
He is.
Then maybe he can answer the question. I’m interested in the answer. I haven’t heard similar complaints about maintainability from the GCC community and they weren’t addressed in the video on the subject.
It sort of sounds like the questions about LLVM’s future. As its relatively a new project mostly funded by Apple, there are some people that are a bit nervous that it might not be supported long term for the uses they have in mind. I think this was the reason why OpenBSD went with PCC: they wanted to be in control of the compiler. And maybe what the BSD Dev here is saying is that they feel like they have more control over LLVM’s future than GCC’s. Is that it?
OpenBSD does NOT use PCC.
My mistake, PCC is included in ports and is capable of building the kernel. I thought they were in the process of replacing GCC, but apparently its not a priority for them.
PCC is NOT in OpenBSD’s ports.
Ok, its in the source code..
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.bin/pcc/
Its been while since I lived in the BSD world. I thought everything in the source would also be updatable in the ports, too.
Does it really make a difference, for the sake of commenting on openbsd’s relationship with PCC?
GCC by itself not so much, binaries produced by GCC are not GPLv3 by virality, but everything -using- a GPLv3 or any “TiVo clause”-ized code on the FINAL device most allow the firmware to be tinkered with. So it is not much GCC by itself, but everything around it that make it much more complicated. That said, I have seen OpenBSD on embedded devices, but never FreeBSD. Most of them just use Linux.
Do Juniper Routers and Switches ( and the recent SSG series Firewalls ) count as embedded devices? Junos is based on the FreeBSD kernel.
http://en.wikipedia.org/wiki/Junos#Architecture
The hate I’m observing here lies within the GNU camp. And I must say, after being away from OSAlert for a while… I’m shocked. Not that things were all sunshine and roses in the old days. But in this thread, at least, it seems even uglier than in the old days. This new crop of GNU zealots are absolutely hateful. But they don’t even confront people publicly. They just sit in the shadows and silently mod thoughtful posts down.
We don’t seem to have private messaging on OSAlert any more. So the public forum will have to do. I note that you seem to have moved. Last we spoke, I believe you had just been through the earthquake. Hope all is well with you.
-Steve
And you are as usual one of the ugliest, you complain about GNU zelots hating when you are in fact someone who has a long record here on OSAlert of making hateful posts against FSF and Mozilla.
You are exactly that which you complain about, so if OSAlert has become worse since you were away, then certainly even worse now that you are back.
And don’t even start with the ‘modding down’ crap, 52% of your 1923 comment votes have been done to mod a comment down. Again you are truly the problem which you pretend to despise.
Wow. Take a chill pill before you have a brain hemorrhage. It really makes me wonder if all the effort I’ve put in at work to promote Linux has been for the good. It certainly would have been easier and more lucrative to go with the flow and support Windows instead.
If I took your brand of drivel seriously (which I don’t) I’d probably just give in to the pressure and move my hundred users to Windows desktops running against 2008 Server and be done with it.
But I don’t because I realize that the fanatics here on OSAlert just don’t matter in the real world. But watching the passion is absolutely fascinating.
Go forth and froth some more.
-Steve
Ah, but you are the one with a fanatic crusade against FSF and Mozilla, unlike you I don’t have such crusades. Several of your posts in this thread has contained an attack on FSF even though it had nothing to do with the subject (in fact your first post in this thread was nothing but an FSF attack), you are the archetype of a person nurturing fanatic hatred and you have been doing this for as long as I can remember.
Meh for me I come here occasionally, make a post then bugger off for a few days then come back to see if there is any more interesting news that I might be interested in. The fact that people make claims that a project makes a decision based on ‘love’ or ‘hate’ really shows how disconnected they are from these projects and the fact that those making the decision do so on the basis of something pragmatic and reasonable and not emotionally drive.
I’ll add (and feel perfectly free to disagree) that a little idealism injected into the decision-making process is often a good thing, as long as it is balanced against the range of other responsibilities.
It’s no secret that I believe in Linux. (I choose those words carefully. The passion is there.) But I also have responsibilities to my customers, and must be careful not to betray the trust by pounding a round peg into a square hole simply based upon my own enthusiasm.
Very often I have a situation where Windows, or some other proprietary software package, might be a quick fix for this or that. And I have to put on my thinking cap, and sometimes even dream about the problem for a few nights, before I come up with a solution which I truly feel is better for the long term. And I might have to hustle up and give time away for free to make it happen. Because management would not always be interested in my view of the long term, and need my solution to make sense in the short term. But I don’t mind.
And then… there are those times that I’m beaten… I lose… or maybe I’m just too tired to put in the extra work. There are times that I can’t come up with an alternative solution which would not be of a ’round peg in a square hole’ nature. At which point, my unpleasant choice is clear.
It’s not always easy to balance responsibilities well enough that past decisions don’t keep you up at night.
And I suspect that Free OS developers face the same kinds of issues, in this vein, as do admins.
Edited 2012-01-13 17:59 UTC
No, its GPL3 only that people have an issue with. If it was just GPL, they wouldn’t already be using FreeBSD as it has GPL V2 code already in it.
But thanks for the link that was pretty decent. Actually worth the time it took to watch it, which is pretty dang rare for web videos.
So a few companies effectively decide what ends up in the FreeBSD “project”.
Interesting.
Wrong. The developers do.
Edited 2012-01-13 16:18 UTC
First comment brings up the licenses, great. I can already see where this thread will be going. 70+ comments about licensing from non-developers and 5 comments about the actual release.
Yes, the GPL zealots will start posting. That’s a given.
Frankly I couldn’t care less about licenses but maybe someone can answer what I DO care about which is the USERS. I sell boxes and laptops to Suzy the checkout girl and Brian the backhoe operator, you know, normal folks?
Well despite all the claims of “user friendly” Linux distros I have yet to have a single one pass my “is it safe?” test which simulates what my customer would go through with just 3 years of ownership. I take a distro from 3 years ago, install it, make sure ALL the hardware works, then using whatever GUI the system has update it to current. I have YET to have a Linux that didn’t puke on its own drivers and end up with a broken mess that can ONLY be fixed with CLI that is above the skillset of my customers, thus making Linux useless.
So my question is thus: If I take say FreeBSD or PCBSD and give it my “is it safe test” what are the odds it will have 100% functionality? The great XP dieoff is underway and frankly i’m getting insanely overpowered XP boxes dirt cheap, we’re talking 2.3GHz-3.6Ghz P4s with 512Mb to 2Gb of RAM and DVD burners, more than enough power for your average FB user. But unless I can find a free distro that is actually worth using I’m gonna have to can the whole lot, just strip them for parts, because a Win 7 HP license costs more than the box and I can’t find a source of Win 7 Starter for system builders.
So if anybody can answer this for me I’d be grateful, we aren’t talking funny hardware, the same Realtek sound and the big three IGPs everybody has had for ages, you know bog standard Intel and ATI chipsets, plain jane boring. you’d think with well known hardware like that Linux would work but nope, broken city. Is BSD any better? Is it worth me wasting the bandwidth or is it another CLI fest? Remember my customers are NOT geeks, no CLI allowed, it has to be easy peasy clicky clicky. Is BSD up to the challenge or should I not even bother?
In FreeBSD you must set up everything with CLI and the standard way of installing software is compilation.
My recollection is that PC-BSD, based upon FreeBSD, has a more Windows style installation convention. The packages are binary, and include all the needed libraries, which may or may not be the libraries provided by the OS.
I’ve criticised this in the past, for various reasons. It means more files must be replaced from various sources to address security vulnerabilities. And results in greater memory use, overall. But it’s definitely a cut above the Linux model for avoiding library incompatibility problems. I’d say that FreeBSD’s PC_BSD spinoff has Linux beat on this one. At least regarding user-friendliness and future-proofing of their packages.
-Steve
And the problem with that is … ?
You learn a heck of a lot more about your system when you configure it by hand.
If you need hand-holding and pointy-clicky GUIs, there’s PC-BSD (which now allows you to select the DE that’s installed, you’re not locked into KDE like previous releases).
Hi bassbeast – have you tried Linux Mint? It is renowned for its user-friendliness.
I use it and am very happy with it. It has a graphical tool (Synaptic) for installing apps.
I’d definitely recommend it if you haven’t tried it.
Edited 2012-01-13 02:16 UTC
I think the point he is making is that brain dead people with jobs that drain the life out of them aren’t interested in learning anything about an OS. So they buy Windows boxes from him, and the MS ecosystem allows their Win98, WinXP, etc. systems to limp along for about 10 years after the product should have died. There’s definitely money in that, if you know how to extract it.
I don’t think he’s talking about people who would like Mint, else they’d have found it themselves.
Isn’t that a bit harsh? Just because they have other priorities in life, doesn’t make them braindead.
Below a certain level of weekly sleep, self-esteem, and other life comforts, humans reversibly lose a certain amount of intelligence and become closer to expensive programmable machines.
It does not necessarily have to be their fault, or to be about priorities they consciously choose.
Yes I have, I tried my “Is it safe?” test and it puked on its sound and the wireless. Oh and the DE ended up with this weird bug where half of the time any changes in the network manager wouldn’t “stick” past a reboot and ONLY would stick if the changes were done via CLI which is right out.
So far I have tried Ubuntu/Mint, PCLOS, Mepis, Fedora (I know its bleeding edge but I got tired of being told “Use Distro X” so I did it just to prove it would fail) Puppy and one other I can’t recall off the top of my head. basically everything listed as “User friendly” I’ve done tried and watched as it puked on its own drivers and died hard when upgraded/dated to current.
The simple fact is I need a free OS that will last 6-8 years either on a single install WITH security updates (none of this “Linux is magical and doesn’t need security patches” craziness like they tried to sell me on Linux Insider) or it HAS to be able to do in place GUI upgrades without puking and dying. I have tried ALL of the above and not a single one ended up with 100% functionality even when using a lousy 3 years as a metric, much less the 6-8 years that is the average lifetime of a desktop in this area.
Its kinda sad though that here we are in the start of the “great XP dieoff” where millions of frankly insanely overpowered desktops and laptops are being sold for pennies on the dollar and I can’t find a single free OS that will pass muster. I have been told every kind of BS from “Linux doesn’t need security updates” (Lie) to “You should teach your users to embrace the power of CLI” like its the fricking force (not happening as normal users don’t want to be programmers) to “You must be a M$ Ninja!” for daring to point out that the Koolaid is bitter and the support situation is horrible.
The average Windows gets 10-12 years of support (XP-2001-2014, Win 7 -2009-2020) so lets split that in half and say…ohh 6 years. Show me ANY free OS that I can either get 6 years worth of patches WITHOUT upgrading or if I must do the upgrade deathmarch show me one that can get 6 years worth of in place upgrades without puking on its own drivers and committing suicide. My users aren’t gonna learn bash programming or do forum hunts which are beyond their skillset, so if you think FOSS is ready for the desktop here is your chance. i’m giving you a typical home user with typical requirements, provide them with a working solution that will not end up costing MORE than simply buying a new Windows license. I wish you luck because so far nobody has been willing to step up to the plate.
As for me it looks like I’m gonna have to scramble to find someone who will sell me OEM Win 7 Starter or embrace piracy like many of the other shops because I can’t sell machines that are gonna need a wipe and reinstall that may or may not puke on the user’ files and settings every 6 months. Don’t say it can’t either because I have a Linux admin friend that has given up on Linux and is looking at Mac after the last OpenSUSE upgrade puked on her system and wiped out 3 years worth of emails.
Haven’t you heard of Red Hat linux distro?
Did you miss the “doesn’t cost MORE than Windows” part? RHEL costs MORE than Windows, a LOT more. It costs $299 a year unless you choose “self support” aka forum hunts which is $49 a year. windows 7 HP costs on average $89 and gets you updates until 2020.
You see THIS is the problem folks, there simply isn’t a Linux desktop in existence that gives decent QA and reliable updates without costing MUCH more than Windows. There is simply no way to even break even on $140 machines with a $300 a year RHEL license, and if you go CentOS you are right back to forum hunts.
Most home users can NOT tell you the make/model/rev of even one piece in their machines, much less every piece, which means doing forum hunts is right out, and even if you could somehow magically take them to the correct place in the forum (no small task) every single time every time I’ve gone to the forums the code I’ve been given is not only CLI (which means a single mistake and its boned) but it has ALWAYS needed to be “tweaked” because it was written for Rev A/firmware B and you have Rev C/firmware D and if it isn’t tweaked it doesn’t work. this is waaay beyond the skillset of a home user, heck you might as well have told them they need to write their own wireless stack.
Never before in history has Linux had a greater advantage yet once again nobody is gonna step up to the plate so it will pass you by. Users are more than ever using only online apps, the great XP dieoff is placing millions of overpowered desktops and laptops on the market dirt cheap,and MSFT’s current pricing makes Windows Home out of the question, yet most of these machines will end up with pirated Windows or thrown in the trash simply because the community refuses to accept Suzy the checkout girl isn’t a C programmer in disguise. What she needs is the PC equivalent of a toaster, you push the button it goes. With automatic updates and a decent AV you get that with Windows for a decade at a time, with Linux as I’ve seen first you’ll be lucky if it survives the first 6 month upgrade with 80% functionality.
For those that say Linux is ready for the desktop here is your chance to prove it, we are talking HOME users, not geeks nor programmers, they need a functional system that will last a minimum of 6 years with updates/patches or with upgrades that don’t puke on the drivers. these people can NOT do forum hunts, they can NOT use CLI, and the hours of training required to teach them those skills will again push the price higher than the $89 for Windows HP.
I have 4 boxes sitting here with more on the way so i can test your solution and see if it passes a 3 year simulation, so lets hear it, I’m all ears. I’d hate to have to sell these machines to the shop down the road just so he can slap a pirated Win 7 and call it a day, but that is what its looking like I’ll have to do because so far all FOSS solutions have failed one or more requirements.
Never heard of backups?
She HAD backups, that wasn’t the point, the point is that OpenSUSE puked on her system and left her with her only home machine hopelessly broken for over a week! When it was done she had NO network, NO wireless, and nothing but single user mode, that’s it, nothing but a blinking cursor and no way to Google a solution!
Now if someone who makes their living administrating Linux servers can’t even get a desktop to upgrade without it puking what chance does my home users have? I’ve been building machines since before Windows even existed and I have had to spend days doing forums hunts and tweaking fixes to get a system back up after an upgrade puked, what chance does a user have that doesn’t have the skillset to do a step by step troubleshoot on a broken system and who may not have a functional system from which to Google have?
Everyone says “Linux is ready for the desktop” but IME that is ONLY true if someone gives away free lifetime support or if that person is a geek with the skillset required to do serious troubleshooting and administration. Now since i can’t afford to give away free lifetime support and my customers are NOT geeks with the skillset required to be admins where does that leave the 99% of the planet just like them? just because YOU think something is easy does NOT mean the rest of the planet would find that easy. i’d say those with the skills to troubleshoot a broken Linux constitute maybe 0.4% of the population and the Linux numbers reflect that.
Give me a distro that will last as long as Windows with updates, or that can upgrade with the GUI without puking and I’ll be slapping it on tons of machines. but so far I’ve tried every “user friendly” distro and ended up with broken messes. I truly wish it wasn’t so, because I know these boxes will end up being sold to the shop down the street which will just put pirated Win 7 on them, but I can’t sell machines that are gonna be broken in less than a year and I will NOT sell machines with security updates disabled.
Increasingly, Suzy and Brian are discovering that a smart phone and game console will cover most of their needs. Personally, I’d be happy if the AOL 2.0 set would move on, stop calling themselves “geeks” (with a giggle) and sequester themselves on FaceBook and Twitter, allowing the rest of us to get on with our legitimate tech interests.
Edited 2012-01-13 02:48 UTC
If all you did was make calls, text people, watch movies or play videos games that would be true. But the average person does more than that so the comment couldn’t be further from the truth.
So you’re the one handing out all those “geek” badges?
Out of curiosity, what OS are you selling?
Ah, I see. You’re selling Windows.
Edited 2012-01-13 04:38 UTC
I’m not selling any OS and I don’t use Windows.
smartphones have tiny screens, limited CPU/memory, poor input and very limited apps. The input issue is even worse with on screen virtual kb’s.
consoles have poor input and limited apps and some still with limited CPU/memory.
Neither comes even close to truly replacing a workstation or laptop even for the average user.
You can be a geek about other things than computers. I know a lot of people that know far too much about My Little Pony.
I have yet to have a single one pass my “is it safe?” test which simulates what my customer would go through with just 3 years of ownership. I take a distro from 3 years ago, install it, make sure ALL the hardware works, then using whatever GUI the system has update it to current. I have YET to have a Linux that didn’t puke on its own drivers and end up with a broken mess that can ONLY be fixed with CLI that is above the skillset of my customers, thus making Linux useless.
What I’m going to write is probably not what you want to hear, but your “is it safe” test is pure balderdash. You are expecting a Linux or BSD distribution from 3 years ago, which is ancient, to update to current without a glitch. You are asking for a level of backward compatibility that even MS isn’t offering. Try to update an adequately powerful machine with XP to Windows 7 without the appropriate third party software and driver disks handy and see the spectacular ways in which an update from ancient software to current can blow up.
What people seem to forget is that Windows XP is a fluke in the history of consumer Operating Systems. Microsoft never meant for XP to be the defacto OS for over 10 years. This lull in OS upgrades was beneficial for users, but it never was the general rule. The upgrade cycle was supposed to be “Buy a new machine every 3 years”. After Vista we have entered that cycle again. So your “Suzy the checkout girl and Brian the backhoe operator” will have to buy new equipment every 3 years with MS OSes. (Assuming that MS will keep upping the system requirements.)
If you care enough about selling FOSS systems to “Suzy and Brian”, why don’t you treat a Linux or BSD system the way they are meant to be treated? No endless updating to current if that can’t be done by a clueless end user (and they can’t). I’m a Linux user, so don’t ask me about the intricacies of BSD. BSD users are infinitely more qualified to talk about that. What I can offer as advice is a simple one about Linux and it might even give you some auxiliary repeat business.
Set up your machines with an easy to use (NOTE! USE not UPDATE to current) distribution. Disable the dist-upgrade option of that distro, so that the major version is frozen. Put the /home directory on a separate partition. Offer your customers a “refresh package” every couple of years for an appropriate fee and update the machines yourself to the latest and greatest. LTS versions should be a great target for this. Your customers don’t get headaches, you get a bit more business, everybody wins.
Then again, you could also continue your (futile?) quest for the perfect software which never breaks. Your choice.
They aren’t upping the system requirements.
Windows 8 will use less resources than Windows 7. BTW I actually have the platform preview and it does seem to take up less resources than Windows 7.
Windows 7 used less resources than Windows Vista (Vista used 600mb for dwm.exe (aero), 7 doesn’t use anywhere near that much. Also 7 works far better on Dual and Quad Core systems than Vista or XP ever did).
Windows 7 works well on machines from 2005. XP RTM running well on machines from 1995? I don’t think so.
Edited 2012-01-14 23:11 UTC
Lucas, I have registered to your site but I can’t create a new topic, neither I can reply to some other topic.
All other guys, please forgive me, but we don’t have a personal message system on osnews.
IT isn’t my site, I just get drunk and talk bollox on there.
You need to post an intro thread in the new members area to prove you aren’t a tard. Then you are most likely to be granted access.
Actually, there is. In your profile you can send messages to your friends and fans.
It’s a rather glaring deficiency, isn’t it. We used to. I’m not sure why it was dropped.
You comment I’m afraid is balderdash friend. you see I DON’T HAVE TO do that with Windows because it has this little thing called SUPPORT which apparently is an alien word to Linux users. Xp? 14 years? Minimum average lifespan of Windows OSes? 7 years. Win 7? MSFT’s roadmap has it supported until Apr 2020. Show me ONE distro, just one mind you, that will give a solid 7 years of updates with NO need to upgrade and i’ll happily toss my “Is it safe” test because otherwise you HAVE NO CHOICE but to jump on the upgrade deathmarch.
Even what Linux foolishly calls an “LTS” is a lousy 3 years and I’ve found other than a few core packages after about a year updates for anything other than the core files is pretty much gone. Also Linux in another brain dead move often has software that demands kernel y when you have kernel x which means unless you have the skillz to compile from source you can give it up.
I have XP boxes in the field that have been there since 2002 with NO bugs and NO intervention from me other than hardware upgrades. It is THAT I need, not some fiddly PITB where my users are gonna be expected to become bash programmers and do forum dances just to get the stupid thing to run more than 6 months. if you can’t give the common man that you have NO business saying “Linux is ready for the desktop” because that is a LIE. That is why Walmart, MSI, ASUS (who actually started the whole netbook thing with a Linux distro and no longer sells a single Linux unit) has abandoned Linux. Its simple, the QA isn’t there and after sales support will eat through any savings on licenses and make Linux a money LOSER. Dell has to pay an entire dev team to keep their own personal repo which means they lose money on every ubuntu machine sold. i can’t afford to lose money on every sale or hire a dev team friend, so i need something that not only works today but will STILL be working 4 years from now. kinda sad that I can’t even find a single distro that gives even half the support time of a Windows edition, isn’t it?
Oh and I hate to break the news to you but homes users DON’T BUY SUPPORT CONTRACTS so your entire plan is a giant fail. Nobody is gonna want a Linux computer that costs MORE than Windows which is exactly what your “plan” would entail. Windows 7 HP is $87, my time is $35 an hour, do the math. Just paying for a single “refresh’ would put the cost of Linux over the cost of Windows. A wise man on this site I believe said “Linux is free if your time is worthless” but I don’t give away my time so I need a product that when it walks out the door it remains functional with minimal work and skill required of the user. Windows with any non IE browser and a good free AV like Comodo fits that requirement, and the average length of desktop ownership in this area is 6 to 7 years as they get handed down to family. Can you show me a SINGLE distro that will give me that length without having to wipe and reinstall? Which is why I asked about BSD because I can already tell you the answers with regard to linux, and its “not a chance”.
RHEL, CentOS, and Scientific Linux. 7 years of support without the need to upgrade. CentOS and SL require no support contract. And both of those have demonstrated a real 7 years of free support for two or three release cycles of RHEL.
CentOS and SL both supported RHEL 2.1 through RHEL 2.1 EOL. But RHEL was new, and they had some ramping up time. They didn’t release for 18 months or so after the RHEL 2.1 debut.
CentOS and SL 3 was supported the full 7 years. And it saddens me to say that CentOS 4 reaches EOL at the end of next month, after 7 years of reliable service to me, and to the rest of its user-base. (I’ll miss that one!)
RHEL 5 was released on 3/14/2007. CentOS and SL will support their versions for another 2+ years.
RHEL 6 was released 11/10/2010 and will be supported by them until very late 2017.
Your checkout clerk customers don’t even keep their cars that long.
PC-BSD is pretty solid IMO.
A Red Hat EL 6 clone like CentOS or Scientific Linux would be a good fit. Not to go back three years, Read: you should install the latest, but it will be supported until at least 2017 and would be good enough for people who don’t need bleeding edge.
After all, once the media player has been fed with all necessary codecs you only need to update from the repos.
That should be long enough, and you can probably long before that sell them a machine with 7 on it.
It would be up to you as the reseller to enable additional media repos and set up the desktop a bit nice, with a shiny wallpaper maybe, and once your GOLD image is done you’re good for a while, maybe spinning a new one for updates from time to time.
Have you ever tried Debian? Granted my personal experience is likely to be different from yours and other people here, but I have never had a Debian old_stable->stable upgrade fail. I have had plenty of Debian testing/sid upgrades fail, which is part of the fun of running testing/sid, but stable has always served me well.
As my avatar may indicate, I very much like the *BSD’s. That said, I do not think FreeBSD is a good fit for your needs since it is a cli-oriented system that requires a user to read manuals and work directly with the cli to install/patch/upgrade software.
True enough. But then we have PC-BSD which is user friendly.
Look. I voted you up. But that doesn’t mean I agree with you. I just don’t think you should have been modded into oblivion.
Certainly, I’ve been pissed off enough by FSF arrogance in the past to seek to have an FSF-free home system. It’s not easy. That small percentage of code needed on a Linux system which derives from FSF is hard to replace. And most people in the Linux world are pragmatic enough not to try just for spite’s sake. The situation is somewhat better for the *BSDs. At least FSF doesn’t have a strangle-hold on their libc.
It makes me feel better to have the option to jump ship completely, if I choose. The FSF is annoying… and sometimes downright scary. And Linux distros had better hustle to stay ahead. Someday, we might just get *too* complacent. (As opposed to the “short of the limit” level of complacency we currently maintain re: the other unix-like OSes.) And our hard-working competitors would have the well-deserved opening they’ve understandably been waiting for.
As a Linux guy, I’m quite happy for my *BSD counterparts.
-Steve
No, it’s just a different philosophy that isn’t compatible with the GPL. While there certainly are people who do hate it for most it has nothing to do with hating.
It’s like saying Linux hate non-GPL.
Edited 2012-01-13 04:52 UTC
Exactly, preferring something doesn’t mean you have to dislike or hate something else. BSD and GPL cater to different preferences, sometimes those preferences change depending on the type of code in question.
I think it’s time I give FreeBSD a try.
Same, its got ZFS version 28 support which caught my eye… I’m still running OpenSolaris SNV_134 which has ZFS 22.
Anyone experienced with ZFS on FreeBSD want to comment on its stability?
I used it for quite some time now, and had no issues at all. Which does not mean that there are any, if you want to be sure, take a look at the mailing list archives: http://lists.freebsd.org/pipermail/freebsd-fs/
Unfortunately ZFS version 28 is the last one released by Oracle to the public. Later versions will not be available, so the good stuff like encryption will probably not come to FreeBSD for quite some time.
Lot of original ZFS developers has joined Illumos/OpenIndiana. Even the head architect and creator of ZFS have quit Oracle. These developers from Sun have already developed ZFS functionality that Oracle does not have. The FreeBSD developers does not know how to develop ZFS functionality, they just port it from the original ZFS developers.
So if you run OpenSolaris, you can easily upgrade to OpenIndiana via IPS package systems, and then you can dualboot via snapshots in GRUB.
Cool, I just realised it’s not that hard to go from Osol to Oi:
http://wiki.openindiana.org/oi/Upgrading+from+OpenSolaris
Will try that first and see how FreeBSD goes
I’m not sure if “does not know how to” is a fair summary – after digging around in that code for as long as they have, I suspect pjd & friends are capable of adding things to it. It’s more that they have so far wanted to stay close to the openSolaris versions so that patches and upgrades for it will apply as cleanly as possible; they only have a limited supply of man-hours, so there’s no need to make it harder than necessary.
We’ll see where they go in the future. It’s quite plausible that they’ll start tracking the ZFS version in openIndiana, instead; I can’t remember them saying anything about it, but I don’t follow the relevant mailing lists either (except -current).
I’ve been running FreeBSD 8.1 (I think) for about a year and ZFS has been rock solid.
I’ve also run ZFS on Nexenta and OpenSolaris, and quite honestly, I prefer FreeBSD as I’ve not experienced any performance issue with ZFS but also I get the added bonus of a more preferable (for me) OS.
I’d definitely recommend it.
We’ve been using ZFS since it hit 7-STABLE way back when. Other than some growing pains around memory usage and poor vdev setup (don’t do a single 24-disk raidz2!!), things have been very stable.
We just put into production 2 FreeBSD 9.0 storage boxes (one upgraded from 8.2, the other clean install of 9.0) using raidz2, lzjb compression, and dedupe. These are backups boxes running our custom rsync scripts to backup over 150 servers every night, creating snapshots every morning, and pushing the data to an offsite server during the day.
The things to note for using ZFS on FreeBSD:
– use 64-bit hardware and version of FreeBSD
– stick as much RAM as you possibly can into it
– use the fastest drives you can afford (avoid Green drives)
– be sure to stick an SSD in the system for use as L2ARC (read cache)
– if you can afford a fast SSD, use it for SLOG (write cache)
If you are going to play with dedupe, you need at least 16 GB of RAM, preferably more. And your RAM needs will only increase as you add data to the pool (the rough calculation is about 1 GB of RAM per TB of deduped data in the pool). You can use an L2ARC device to lower the RAM requirements, but that only helps so much.
If you want performance, go with mirror vdevs. If you want storage space, go with raidz2 (or even raidz3) vdevs with small numbers of drives per vdev (keep it under 10). The more vdevs in the pool, the better the overall performance of the pool (meaning a 24-disk pool using 1 raidz2 vdev will be much slower than a 24-disk pool with 4x 6-drive raidz2 vdevs).
It may take a little tuning of /boot/loader.conf settings to make things perfectly stable, although the auto-tuning in 9.0 has improved a lot. All we set now is the vfs.zfs.arc_max to about 80% of RAM.
intel HD gfx are not supported very well (uses VESA mode..) and suspend/resume to ram is problematic.. As I read this will be fixed on/after 9.1 …
GPL has its advantages. But it really is a tragedy that we can’t (or won’t) give them our driver code. Lot’s of work there that will need to be duplicated. Of course, there is a lot more to porting such a driver than getting over the licensing hurdles. Especially considering the graphics memory management differences. Sun (hypothetically) changing their licence on ZFS would likely have made no difference to Linux’s collection of filesystems. But still, it’s a damn shame. These guys are our allies. And we have to refuse to give them code because that would make it available to our enemies.
I’m not about to get into a long GPL vs BSD flame war. And I have no answers. But I *will* say that the current state of affairs is a dirty shame. We should at least all acknowledge that.
Edited 2012-01-13 03:34 UTC
The license used on the code isn’t the issue and the code in question is not GPL’d.
Edited 2012-01-13 03:41 UTC
Yes it is. GEM is integral to any recent version of the Xorg intel driver. Try running an Intel graphics chipset without it. I have an Intel graphics chipset. So I know this very well.
I was talking about the X DDX driver (xf86-video-intel) and Mesa. You’re talking about the Linux kernel code. It’s of no use to anyone else besides Linux so the license is irrelevant.
The *BSD OS’s are already using the Intel driver with GEM without any GPL’d code.
Yeah. Whatever you say.
Can’t accept reality I see.
http://www.openbsd.org/cgi-bin/cvsweb/src/sys/dev/pci/drm/
No GPL in sight.
*nod* Same goes for the DDX driver (xf86-video-intel), Mesa and all of the rest of Xorg.
The FreeBSD Foundation is funding work on “porting” kernel mode setting (KMS), GEM, and other bits to make newer Intel graphics work. There’s beta-quality bits available for testing already, with several users reporting success using them.
And there’s even someone working on the KMS stuff for the Radeon driver.
That’s really good news. It means that if we will have KMS in FreBSD than Wayland is on his way and then we will say bye to X.
I presume you mean that it’s a good thing for smart phones. Desktops are much better off with Xorg than with a smart phone display layer.
I’m migrating ~100 users away from Ubuntu over Unity, the smart-phonetization of that distro, and last but definitely not least, the lack of the network transparency we require. (We use NX/FreeNX extensively for our local Windows terminals and branch offices.) We need to upgrade, and my users would skin me alive for taking them to Unity or the mostly nonfunctional gnome2 emulation shell being offered.
And with KDE and Gnome devs having lost their minds, as well, the only reasonable haven is CentOS, until this mania blows over.
If you choose to move to a Wayland desktop, you are beyond the help of any advice I can give you.
Well I’ve always thought BSD or LGPL is the ‘best’ licence for ‘driver/library/framework’-type code, but then again that’s just my personal preference.
At the end of the day it’s up to the actual programmers writing the code, or those paying them to write it to decide which licence to use.
How uncharacteristically reasonable of you. I agree.
I respect the author’s right to choose whatever license he or she deems appropriate. Even a proprietary one.
As a user I certainly prefer GPL or LGPL. But I’m perfectly cool with any Free license, even if some might find it too permissive.
I find myself caring less about the details of the Free licence, and more about whether the author is treating the OSes I use as second class citizens or not. Hence my concerns about Mozilla, which you referenced earlier. Having been a Unix/Linux guy since 1988, for both desktop and server, I’ve had about enough of being treated as a second class citizen.
That’s why I am particularly scornful when I see users and devs of kindred OSes, like FreeBSD, viewed as somehow being second class.
I’m quite happy to see FreeBSD doing well. Such sentiments seem to upset some people. And I’m not entirely sure why. It doesn’t make much sense.
That has always been my stance on the subject, as you can easily see by any other posts I’ve made on the subject.
‘Concerns’.. lol. Also as I recall it the big gripe you had (atleast only one I’ve seen) with Mozilla was that you hated how they wouldn’t allow distro versions of Firefox to use the name Firefox (which to me makes perfect sense), then you gave your support to Chrome which did the exact same thing (Chrome – Chromium) indicating that whatever Mozilla did to create that huge animosity you have against them likely wasn’t that.
Certainly not me, my favourite (and hopefully future day to day OS) is Haiku and I’ve never understood why some people despise it, sometimes for what seems like nothing but the fact that it exists.
This is more or less spilled milk over the dam at this point. But the way the newly formed Mozilla Corp handled the situation with Debian was pretty cheesy. They didn’t really even give Debian time to do anything. It was more or less “Cease and Desist” right off the bat. Likewise, Mozilla’s cavalier treatment of the very reasonable objections of the Firebird Database devs, whose project name Mozilla stole, left something to be desired. And then there is the continued treatment of non-Windows OSes as 2nd class citizens.
Of course, there are no doubt some number of other lower profile examples I won’t bother to go into. It’s really not worth the research time.
In the first 2 cases, Mozilla acted wrongly. In the 3rd, it’s simply a matter of not wanting to support a project that doesn’t care that much about the OSes I use at home and at work. Especially when the Chrome/Chromium devs view Linux as their premier platform.
I do believe in calling a spade a spade. I think you are mistaking that for some sort of vendetta. I don’t waste time fretting about such things, these days.
I am, however, perceiving that you harbor some sort of vendetta against me, personally. Unfortunate. But I don’t fret over that kind of thing, either.
It was ‘rename your unofficial version of Firefox to something else’, exactly what Google tell people who would try to distribute unofficial versions under the Chrome trademark.
The reason they enforce this should be obvious.
Of this I know nothing.
Are you claiming Chrome (which you endorsed) was treating Linux as a 1st class citizen? How long did it take for them to even get a Linux port out? Even longer to get it up to par with the Windows version.
I remember you well from all the vitriol you’ve spread here on OSAlert over the years, and seeing you back from your hiatus the first post I see is one where you’re back to your usual FSF hate fest, picking up right where you left off. So yes, I think you are nurturing the same old ‘vendetta’ of which we’ve seen countless examples of in the past. Unfortunately.
I don’t know you personally.
license war aside. I think FreeBSD trades blows for features and compatibility with any other OS out there. all it lacks is the performance lead you get from optimization.
most benchmarks will show FreeBSD stuff lagging behind par. but all in, this is still a really crazy and wonderful place for such software to be. and it is lead by a small organization.
I remember plenty of times that the FreeBSD guys have shown our beloved Linux kernel up for performance. And our Linux kernel devs have had to scramble to close the gap. (IIRC, FreeBSD 7 was a good example.)
Although I disagree with that particular point in your post, I have to wonder why anyone would mod it down. I get the impression that the level of irrational vitriol (as opposed to… err… rational vitriol) on the OSAlert forums is even greater than I remember it to be. And I would not have thought that possible.
I’m wondering if I should stay or go elsewhere. But you know how it is with watching train wrecks in progress.
Can I ask what “more ideologically developped” is supposed to mean in the context of the article ?
I think it means that some folks think that software licensing can be used as an effective political weapon.
It means someone don’t know what they’re talking about.
Might want to also quote the bit that followed.
It means that the BSDs tend to be more conservative, developing along the lines of what they believe is the best long-term option to deliver a smooth developing path – even if that means not always having the latest and greatest features. Linux, on the other hand, takes a more pragmatic approach, and will happily settle for a less ideal option just to get the latest and greatest feature, even if that means having to do it all again shortly after. This is pretty basic stuff.
It had nothing to do with licenses. Settle down, people. Dear lord.
Oh I agree I just don’t know if “ideologically developed” is the right term. It just sounds wrong.
“Idealistic” might have been the term Thom was looking for.
Wonderful that we have another updated, modern OS to play with !!
This is what drives progress, new releases of operating systems that forces competing OS:es to push their development further ……..
We should all be glad that there are many initiatives on the market for OS development, no matter if their in to different philosphies regarding licensing etc. etc.
Remember the 90’s when there were Win 9x and NT….. and not much more…..
Remember the 90’s when there were Win 9x and NT….. and not much more…..
You must be young. What about the BSD’s (veritably old in terms of existence) Linux, OS/2, Amiga Workbench, Mac OS, DR-DOS, etc. Not widely in use is not the same as “and not much more”.
FreeBSD’s development start in 1993: http://en.wikipedia.org/wiki/FreeBSD#History
NetBSD, Slackware, and Debian were also started in 1993 and OpenBSD was started in 1994. In comparison, the Windows “9x” line started with Windows 95, which was released in August of 1995. The 90’s had far more options than just Windows.
BSD Originated in the 70s at Berkeley. That’s why it’s called BSD. It is a real UNIX, FreeBSD was developed because ATT owned a lot of the code in the OS.
http://en.wikipedia.org/wiki/Berkeley_Software_Distribution
Not quite. BSD-Lite and BSD-Lite2 were developed because AT&T owned the copyright to 6 files in BSD.
BSD did not run on the Intel x86 architecture, so 386BSD was developed by Bill Jolie. However, there was a big hullaballoo around the incorporation of third-party patches into 386BSD, and users had to download these huge patchsets in order to make it work on their machines. Out of that mess came NetBSD and FreeBSD (started independently of each other, both based on 386BSD).
Shortly after NetBSD and FreeBSD were launched, BSD-Lite2 was released. FreeBSD then rebased their code on BSD-Lite2. Not sure when/if NetBSD did the same.
And then a little while after that, Theo had a falling out with the NetBSD core devs, and OpenBSD was born.
And many years after that, Matt had a bit of a falling out with the FreeBSD core devs, and DragonflyBSD was born.
I just dropped by to say that I am excited that FreeBSD is doing so well. A few years ago I was worried that Linux would eclipse it and it would die off. From my perspective, it seems stronger than ever.
ZFS is pretty enticing and I have wondered if perhaps I should not switch over to using FreeBSD for software development. I wish that VirtualBox ran on FreeBSD. I need to do more investigation into how well running virtual machines on top of FreeBSD works.
Perhaps we will see FreeBSD on a tablet or smartphone someday soon. iOS* does not count.
As it happens, I am posting this from OS X (which is a cousin of FreeBSD). I am really a Linux guy though and use it almost exclusively (Debian or Ubuntu on the desktop and RHEL-alikes on the server). I also use Windows from time to time. I probably waste as much time on Haiku as I do real work on Windows though (which is all I would use Windows for).
* Although I have an iPhone, IOS will always be Cisco’s Internetworking Operating System. I get confused all the time when I see references to iOS (by Apple).
VirtualBox OSE does run in FreeBSD, quite nicely actually.
http://www.freebsd.org/doc/handbook/virtualization-host.html
I typically have an instance running a Windows XP image on 9-CURRENT (haven’t updated yet since -RELEASE), so that I can run Visio, and get full printing and scanning capabilities from my Brother laser MFC that doesn’t have a proper CUPS and SANE driver. There’s even USB pass-through nowadays, although it’s not perfect (not sure if it’s any better on Linux or just the same).
VirtualBox does run on FreeBSD and runs a great deal more stable than when I’ve been running VBox on other platforms (I’ve had up to 5 concurrent VMs running with uptimes of months without any issues)
To be honest, I doubt that. Not because FreeBSD isn’t capable, but more because that arena is already over crowded so I can’t envisage FreeBSD being utilised that way.
But maybe someone will prove me wrong
You mean Infogear’s iPhone now runs Cisco’s Internetworking OS?
http://en.wikipedia.org/wiki/Infogear
While it is undeniable that here the most important part of this release is a GPL/BSD argument because this is an internet sport by people who have nothing to do with their time, I think that placing too much priority on making BSD-replacements for the base system when there is no good third-party replacement already is misguided at best, if not outright myopic. I care about having a well maintainable and stable system, which BSD is known for, not an operating system which will be able to used by companies as they wish without giving anything back, and making them being able to do so a *priority*. gcc >=4.3 is still very used by companies, and license requirements aren’t really as important to companies as people make them to be. Think of it most companies infringe the GPL and nobody ever goes legally after them.
I used to like FreeBSD once, but nowadays I can’t help but wonder if the FreeBSD and the OpenBSD projects really care the usability of their operating systems or not. One is stuck about this “let’s BSD-everything” and “let’s change everything”… FreeBSD used to have a very good installer but the one in 9.0 is a real step backwards despite them saying it is a sort of advancement while actually it is not for someone not new to FreeBSD. The other one, OpenBSD, is stuck on making you upgrade every six months (which even for an home user is not viable – I do not live for computers or on my computer, thanks), the same BSD-ization craze and on an unreasonably ancient userland.
I do, however, appreciate the reluctancy OpenBSD devs have for signing NDAs, at least they’re being more consistent than the FreeBSD devs which will happily sign NDAs “as long as it is BSD licensed!”
If you want BSD, the best stuff is NetBSD and DragonFly. If FreeBSD and OpenBSD are stuck at gcc 4.2 for “problems”, then why I’m happily using NetBSD-current on the very machine I’m writing this on which ships with gcc 4.5 with no problems at all.
Despite what it might NetBSD-current is what you must try, much more updated than the stable version. It can even run Opera 11 correctly under the new linux emulation (guess what browser I am using?).
As it often goes, what people use the least (NetBSD and DragonFly) are better than what people use the most.
Are you sure you’re running GCC from base? The CVS log for src/gnu/dist/gcc4 says it’s GCC 4.1, maybe you have 4.5 from pkgsrc.
Edited 2012-01-13 10:03 UTC
I find OpenBSD more usable for some things (BGP, Firewall etc) than Linux. So obviously it has usability. Maybe not for you but that’s not a universal measurement.
You don’t have to upgrade every 6 months but it’s recommended. If you have an issue with this then OpenBSD is obviously not for you but please understand that not everything in this world is designed to cater to your tastes.
Unsubstantiated flattery will get you nowhere.
What ancient userland?
Edited 2012-01-13 11:42 UTC
You’re talking about OpenBSD, of course. But this prompted me to look over FreeBSD’s release history. It looks like they target about 2 yr release cycles for major versions. And maybe 8-12 months for minor versions. I’ve always felt that the Linux world could benefit from longer cycles. This makes FreeBSD look more interesting to me.
And each major release looks to have an impressive ~5yr support cycle. It looks like you do have to upgrade through the minor releases to get that. But I assume their idea of a minor release is similar to RHEL’s. i.e. very conservative, no big deal, just set it to go automatically with confidence that nothing will break.
It certainly appears to have some attractive features for business use. After supporting Linux for 16 years, I’ve become tired of the constant, often pointless churn, and have moved back to CentOS pretty much everywhere. (That is to say in those places that I ever decided to diverge from it at all.) I do have a couple of upgrades to do in the next month, since CentOS 4’s 7 year life cycle is up at the end of February. And I must say, I’m going to miss that release. It’s had a fantastic, and virtually flawless run.
Edited 2012-01-13 17:35 UTC
sysinstall only supports MBR partitioning with BSD labels. Meaning you cannot use disks over 2 TB. There’s also no support for any of the nice GEOM classes (gstripe, gmirror, graid3, graid5, ggate, hast, etc). There’s no support for ZFS. There’s not even any support for labelling filesystems, partitions, or disks.
bsdinstall supports all of the above.
sysinstall had a bunch of crappy post-install configuration features that were half-assed at best, and cause more issues long-term than they fix in the sort-term.
bsdinstall has none of that.
sysinstall is built using a very ancient, highly customised version of libdialog that nothing else in the base OS supports.
bsdinstall is built using a modern, maintained version of libdialog that is also used by other software in the base OS, and is the basis of the OPTIONS framework of the ports tree.
You have to download a specific version of the sysinstall-based installation CD in order to get a LiveCD where you can do troubleshooting or customise the install.
Every bsdinstall CD is a LiveCD, and you can drop to a full-functioning shell at various parts of the install process so that you can manually do things that aren’t (yet) supported by the TUI.
sysinstall is dead. It’s been on life-support for about 15 years longer than it should have. It’s time to let it go.
bsdinstall is not perfect. But it’s a hell of a lot better than sysinstall ever was.
And … that’s different from Ubuntu how? Or Fedora? Or any of the other Linux distros with a 6-month release cycle?
Lets talk about FEATURES shall we? (not LICENSES)
But… but… there’s “BSD” in the title !
Yep agreed and even if it turned off by default, this is very interesting that Capsicum is integrated into the kernel: it is the first time that a widely available OS provides capabilities http://en.wikipedia.org/wiki/Capability-based_security
Note that this feature is different from POSIX capabilities and for example allow Chrome developers to implement their sandbox much more simply.
Simplicity is good for security..
Agreed. As a linux user, I really envy the ZFS filesystem to the FreeBSD users (oh, I’m afraid we’re still talking about license issues here… )
Whenever I watch demos, what I really find myself envying ZFS users are the slick command line tools that very simply and easily manage the partitioning, the lvm, the raid layer, and the fs, all at the same time.
There’s no reason that Linux’s current stack couldn’t have such an elegant solution.
At this point, I usually get recommendations: Oh! Just apt-get this or that. But I’ve never found anything as elegant as what I see in ZFS.
I’ve not been keeping as close an eye on btrfs as maybe I should. It might have such tools. Or might get them. It seems to lend itself more to such an integrated approach, concerns about “rampant layering violations” aside.
Edited 2012-01-13 18:12 UTC
http://bit.ly/ytTJHd
I’m not saying it’s as good as ZFS feauture-wise (because it isn’t) but it does highlight one thing I really like about the BSD’s (and OpenBSD in particular): there’s an integration and coherence in tools and configuration files that isn’t quite there in Linux.
Edited 2012-01-14 10:50 UTC
Just curious: why do people insist on using URL obfuscation (aka shortener) services on websites, instead of just posting the complete URL? Is it really so hard to just paste:
http://www.openbsd.org/cgi-bin/man.cgi?query=bioctl
At least then we can see what you are talking about, instead of having to blindly click the link, hoping it’s not taking us to some horrible cracker site or rick-rolling us.
I understand the use of URL obfuscation services in print, as you have to manually type the URL into the web browser, so the fewer characters the better. But on a website? Where the pasted URL is clickable? What’s the point of obfuscating it?
Edited 2012-01-16 17:55 UTC
What’s the difference from a end-user perspective between this and the popular Linux distributions?
I’ve used Linux a lot and would be interested to know how BSD differs since I’ve never really tried it. Speed? Availability of software?
What would make one choose this over the Linuxes?
It depends on what kind of end user you are, if you are interested in desktops then I think Linux will beat bad in terms of compatible hardware, available software and ease of use with more communities and an easier UX. PC-BSD is ok as my personal opinion is that it’s the best of the BSDs for desktops, however it’s still quite as usable as say mint or Mandriva. I wouldn’t put Mac OSX in the BSD group as Mac OSX tends to only use the user land and not so much of the core BSD technologies such as the kernel.
If however you were a server administrator then I would say that BSD and in particular FreeBSD has plenty to offer. It’s well optimised and performs well under pressure as well as being pretty easy to scale out. The packages/ports are usually well tested and generally once a BSD server goes up its rare to have it come down. It’s been the backbone of a lot of big organisations such as amazon for a while. Again my own personal opinion but FreeBSD tends to approach the server market like the old big iron and Solaris unixs, that everything is carefully planned and nothing is bleeding edge for the sake of it, it’s a good trustworthy server os.
I’ve run a few servers for schools, Opensuse, Ubuntu, Debian and FreeBSD and I liked FreeBSD the best found it stable, secure and fast enough. The documentation worked.
I plan to try FreeBSD with everything installed from ports as a desktop PC just for fun and see how quick and stable it is – when I have time.
I suppose the difference between the two (apart from licensing and the ideological) would be Linux seems more creative and anarchical (not being derogatory) and FreeBSD more planned and conservative.
How many users here are back on Windows?
It seems that the desktop usability will always be an issue to BSD and Linux…
Not that it has anything to do at all with this news item but…..not me.
It seems you’re overly fond of making inflammatory comments.
Well, then this counts only 1 person in OSAlert. There is the second guy below. I myself didn’t yet. But I’m considering. Overall, Thom’s primary OS feels like it is Windows.
My claims are supported with the climbing high skies of CentOS in Distrowatch.com ranking.
Edited 2012-01-13 17:36 UTC
I absolutely cannot imagine it. And I’m not just using that as the common expression that it is. I literally cannot imagine myself using Windows other than while doing the bit of Windows support that I sometimes can’t get out of.
Then again, my historical desktop path looks like this:
Xenix 286/386->Unix 386 3.2->SCO Unix 3.2.4.x->SCO Open Server 5->Linux
starting from about 1988, with my transition to Linux occurring around 1996.
There’s just way too much of an impedance mismatch between the way I work and the way Windows tries to make me work for me to be happy with it.
A switch to Windows seems like a form of hell on earth to my sensibilities. Linux sometimes has usability issues, but they are generally fixable. Windows has usability issues for me which run to its very core and can’t be fixed.
I see. You are more reasonable. However, desktop usability is really becoming an issue in Unix world (as far as desktop usage is concerned).
I agree. To the extent that Linux has a desktop presence. One tends to get a skewed sense of reality when immersed in the OSAlert noosphere. (Oops. I used an ESRism.)
I included this link in a post I made a little while ago. While it’s handy, I may as well use it here, too:
http://gs.statcounter.com/#os-ww-yearly-2008-2012
In a nutshell, *if* we want more people using Linux on the desktop (and that is not necessarily a given) then we need to be doing something differently than we have been.
-Steve
P.S. Same goes for the FreeBSD and PC-BSD guys, too. They face the same questions and the same problems.
Edited 2012-01-13 19:23 UTC
BSD doesn’t face the same problems since BSD projects (FreeBSD, NetBSD, OpenBSD etc) are being run by teams which stand behind their projects. BSD isn’t an anarchy, unlike Linux.
FreeBSD is an operating system. OpenBSD is an operating system. Ubuntu is a distro, Fedora is a distro. Linux is what?
As should be pretty obvious to you, the meaning of the term “Linux” is dependent upon context, as are so many terms in most natural languages. In this context, it is obviously the summation of systems using a Linux kernel. It is both convenient and proper to use the term “Linux” to refer to that set. Please don’t try to confuse the issue with all that silly Stallmanesque “Linux is just a kernel” rhetoric. It is both irrelevant and unbecoming.
Anyway, please refer to the linked statistics. If Linux is at 0.75% of desktop usage, none of the *BSDs register at all.
That’s not intended to be a jab at the *BSD distros. But it’s something I must point out in order to properly address your post.
The *BSDs face many of the same problems that Linux does. But obviously, the two sets are not absolutely identical. No doubt the *BSD’s are doing some things right that the Linux world, by and large, is doing wrong. But the Linux strategy, as a whole, has apparently been more effective at conquering the desktop. Though it should be fairly obvious by now that that is not saying much.
Edited 2012-01-13 20:47 UTC
Ubuntu is an operating systems, Fedora is an operating system. Linux is kernel.
Not that anyone will read this comment way down here, but it’s a great shame that about 95% of this topics discussion has been bickering about licences.
There’s a lot new in this release: not least a huge upgrade to ZFS.
It’s just a great pity people seem more interested in the licensing than the technology itself
*Shock!* *Horror!* And on OSAlert, no less!
Yet some of us still read this.
Indeed, at this rate FreeBSD will be what OpenSolaris could have been.
I can’t wait to have a decent NAS Server at home with ZFS. It will probably be a Mac Mini with a Thunderbolt attached set of 6-8 3TB WD Caviar Green (RAIDZ+Spare) and two SSDs, if the FreeBSD ZFS implementation supports L2ARC acceleration using Flash storage and a second Thunderbolt attached 1Gbps NIC.
I like my storage to be fast and to be able to do the following:
* Be able to keep my VMWare Virtual Machines stored over iSCSI exported ZVols.
* Be able to keep all my files and time machine backups safely with snapshotting.
L2ARC (cache) devices have been fully supported in FreeBSD’s version of ZFS since it was first imported into 7-STABLE. Same with SLOG (log) devices; although it’s only with FreeBSD 8.2-STABLE/9.0 (ZFSv28) that your pool could survive the death of a SLOG.
If you want your storage to be fast, then avoid the WD Green drives. They are, to be blunt, absolute crap. With the first generation of the WD Green drives, you could disable the 8 second head parking “feature” using their wdidle3.exe tool. They’ve since removed that capability from their firmware. WD Green + raidz = worse performance than a single IDE drive.
If you need low-power drives, then look at the Hitachi or Seagate LP drives.
iSCSI works, but using NFS gives you so many benefits. Even VMWare is moving away from iSCSI and other SAN technologies and recommending NFS (and their VMFS on top) for high-performance VM storage. And NFS works better with ZFS than iSCSI.
I would imagine *if* BSD were the first, before Linus published his kernel. I think that all things would have been different. No GNU, no GPL, no Linux, but just BSD and a true Unix.
It was. By 14 years.
Meh, it already happened. It was called SCO. Nobody cared.
I’ve been using FreeBSD since 2.2.5 or 2.2.6, which is going back to the mid 90s sometime. My previous experience had been on Digital Unix, followed by Linux.
I’ve always liked FreeBSD better than Linux. Take this for what it’s worth, one man’s opinion. FreeBSD has always been much cleaner and spartan under the hood. Back in the day there was a demonstrable performance edge with FreeBSD, although that is probably no longer the case.
FreeBSD is for someone who will appreciate a simple, uncluttered unix. It’s also an os that will require you to roll up your sleeves and make it the way you want it to be, which is less true of Linux nowadays with all the work that’s gone into the install and user experience in the various distros.
I hope there is always a FreeBSD. It’s reassuring to have a constant, simple, rock solid unix variant with a single vision of where its going. It’s kind of like an old friend.
As the old saying goes:
FreeBSD is for those who like UNIX.
Linux is for those who hate Windows.