The next major release of FreeBSD, version 7, is one of the most significant so far, with amount of new technologies and improvement largest since introduction of 5.0. Since constantly searching the mailing lists for important changes can be a bit tedious, this page lists some of the more interesting new things in one place.
What’s Cooking for FreeBSD 7?
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
I recently upgraded my system to a better one, and it has a ULI chipset for Sata subsystem. And guess what, I can’t use the FreeBSD 6.2 with my system, where Linux runs like a charm.
FreeBSD needs to support more hardware if it wants to have more publicity.
Just my 2 cents.
>FreeBSD needs to support more hardware if it wants to have more publicity.
If Linux is okay for you, stay with it. Never change a running system.
ULI & FreeBSD:
http://groups.google.com/group/mailing.freebsd.stable/browse_thread…
ULI & Linux:
http://linuxmafia.com/faq/Hardware/sata.html
So in the end, SATA is really new for both operating systems and ULI is somewhat exotic.
Ask the hardware industry first, to produce proper hardware, to offer proper documentation!
Am I wrong thinking that if an Open Source operating system supports a chipset, where another Open Source operating system should support it too?
Both FreeBSD and Linux are open source systems. If one thing is implemented in one of them, it could be ported to another one with a different implementation.
Yes it could, but there are differences:
1) the very liberal BSD license and the GPL with copyleft; it’s a big NO in the system because of the GPL restrictions
2) a different development model. Linux folk tends to work according to “just do it”, whereas BSD folk adheres to “quality first”.
3) last not least the lack of manpower. Almost 50-60% of the patches in Linux kernel are done by big companies[1] and there it’s just the kernel. BSD is always the whole operating system to care about, so developers aren’t taking care for the kernel only.
[1]<a href=”http://lwn.net/Articles/222773/“>lwn.net
“Either way, the results come out about the same: at least 65% of the code which went into 2.6.20 was created by people working for companies.” (fulltime)
Well, reading the comments on the survey, it appears that the data was collected solely on the basis of email address, and that it overlooked some aliases. In short, it’s raw data.
A similar survey a few years ago credited most of the work that I’ve done to other people (using the scientific method of grep’ing the sources looking for email addresses, full names, etc ;-). There’s no evidence that this survey is any better. Doing an accurate survey takes a lot of time.
For the record, the earlier survey was at this URL
http://orbiten.org/ofss/01.htm
No, you’re not wrong. Franck and Oliver failed to understand when you mean. You didn’t ask for FreeBSD to have perfect HW support, you asked for it to have HW support as good as Linux, which is a lot different, and legitimate, as most Linux drivers are open-source (not talking about closed-source drivers), so you just have to port them to FreeBSD.
In a perfect world, you would search on the Internet to see if the computer of your dream is fully supported by your operating system. In the real world it’s not that easy, even if there’s a HW compatibility list on the FreeBSD web site, you’re gonna have a *really* hard time understanding the code for each piece of HW (ie: sk0, rl0, anyone?). Then you’ll also have to face different versions of the same HW code, which you only discover buying the HW. In other words, only seasoned users/devs can tell for sure a computer is going to run with FreeBSD.
But as a rule of thumb, the higher the price of your HW, the more chances you have to have compatible HW 😉
I understand perfectly, but what you say is simply not valid for any particular snapshot in time. Over the long haul, sure, drivers will be ported. But people are only interested at the particular time when they want the support. That driver, particularly for newer hardware, may or may not exist. You have to check, and yes, that can take some work.
The 64-bit nVidia driver is not available on FreeBSD. Why? The kernel needs changes, and these are happening slowly. Open had great Wifi support before anyone else. Now that has diffused to other systems. FreeBSD will have ZFS before Linux. That’s a license issue (and not hardware, but it is related).
I have never had one problem with FreeBSD hardware support over eight different computers, not all of which are x86 architecture. Sure, I understand what rl0 (*brrr*), fxp0 amd em0 mean. The reason is two-fold: my systems all are a few years old, and I looked to ensure that the various devices were supported without resorting to unwarranted assumptions.
“so you just have to port them to FreeBSD. ”
Oh yes, it is so simple. Never mind that you have to make it work in an entirely different kernel design and that you cant use the (L)GPL’d code due to the licensing. Also, it takes no time and resources at all and can be done instantly the moment new hardware arrives.
Am I wrong thinking that if an Open Source operating system supports a chipset, where another Open Source operating system should support it too?
Under the hood, BSD has quite a few differenes from Linux (which is why the binary compatibility thing exists). Due to this, porting drivers can, in some cases, be easier said than done. When it is possible, source-code-licensing issues and not-invented-here pride occasionally get in the way. A lot of BSD folks don’t like the GPL (to put it politely), and FreeBSD prefers BSD-licensed code to keep things simple, so FreeBSD tries to have as little GPL code as possible (the Ports Collection being a glaring exception, of course).
FreeBSD also tends to be more server-oriented than Linux, and its included drivers reflect such. You also have to keep in mind that many Linux distros add additional drivers (especially Mandriva and SuSE). From my experience, FreeBSD usually lags behind Linux in supporting “exotic” hardware, though it isn’t too far behind.
If you want to run FreeBSD, check their hardware compatibility list and choose or build your computer around it. I also learned this the hard way. It’s not as picky as Solaris, but it’s always good to be on the safe side. Also, FreshPorts is a very good place to look for FreeBSD drivers.
Edited 2007-07-12 22:13
FreeBSD also tends to be more server-oriented than Linux
More server oriented? Most of the commercial companies that contribute to the Linux kernel do so with the server in mind. Linux runs on mainframes, FreeBSD does not. Oracle, DB2, SAP, and so on runs natively on Linux, not on FreeBSD.
Linux just happens to have more desktop-oriented contributions in addition to all the server-oriented contributions.
I should have worded that differently, sorry. I meant that FreeBSD is more server-oriented than desktop-oriented, not that Linux was any less server-oriented.
Edited 2007-07-13 17:55
I recently upgraded my system to a better one … [and] I can’t use the FreeBSD 6.2 with my system.
I am amazed by how often people cite this well-worn stupidity. All OSS systems take some time for the latest and greatest new chipsets or Wifi or whatever fully to be supported. If you want to run FreeBSD, for example, check to see that the hardware you want to buy is supported. Most is, but of course there are exceptions.
OTOH, if you want to chose your hardware with very few software compatibility concerns, run XP. Or chose an Apple if their hardware suits your needs. But my g*d, if you spend money on hardware, at least make sure that you can run the software you want.
Sheesh.
Thank you for your polite comment. You are really one of wonderful people here.
Edited 2007-07-12 21:28
Linux doesn’t even support my ethernet card, while FreeBSD does out of the box. Linux doesn’t work with my wireless easily, while with FreeBSD it works out of the box. I think Linux needs to support more hardware…
Don’t take that as me being offensive though, every OS could support more hardware. I’m just pointing out that FreeBSD isn’t horrible on hardware support, and Linux isn’t perfect either.
Edited 2007-07-13 02:39
What make and models are your ethernet and wireless chipsets?
Realtek RTL8101 for PCI Express, which when I tried to install Linux a couple of months ago was not supported whatsoever. This is supported in full in FreeBSD 6.2, and probably earlier.
Wireless is an atheros based card, however the only way I could get the drivers for it were to already have a network connection, and because my ethernet was not supported, this did not work.
I just wish my USB Ralink device (a 2570) would work with Network Manager. Fortunately it sort of works with WiFi Radar. I find it odd that hardware from a company that is very forthcoming with documentation isn’t better supported on Linux.
zfs — good
updated Linux binary support — ok
Security event auditing — nice
no Dtrace — oh well.
All and all 7 is shaping up to be a very big deal, IMO. Keep it rock solid and stable and I’m happy.
This is the “sad” example of copy & paste “news” site.
FreeBSD is a mature system and 7.0 will be a major improvement…but where is the news right now ?
Just remember that FreeBSD is a server first, and has only very recently started to be used on the Desktop more main stream with the availability of PCBSD etc.
And the changes in Version 7, from a ServerAdmins point of view is it’s going to be a VERY good release.
Linux on the other hand has had a lot longer as a Desktop than BSD, and probably has a lot more developers for newer hardware.
Going to be another great release from the BSD team, keep up the awesome job!
ZFS and the Linux binary compatibility update are enough for me.
If 7.0 can mix those planned improvements with the long-standing stability and high-quality traditions of BSD, sure this will be an impressive release.
yeah sounds great, some things are in the new Mac OS X, too. Off course, the uses a FreeBSD like kernel as far as I know. But sounds very good for both systems =)
I recall not so long ago when there was a discussion amongst linux devs that linux needed to catch up to OpenBSD as far as wifi drivers were concerned.
http://hardware.slashdot.org/article.pl?sid=06/06/12/1837238
Complaining about lack of hardware support is retarded. I have better hardware support for my digital camera (a few years old) in FreeBSD than I do with Windows (where kodak decided to not bother fixing their crappy drivers). So what? there are always going to be weird devices that are supported on one OS but not another.
My ubuntu system works better with my webcam than either my BSD or my windows system.
What’s the moral of the story? Different things have different strengths & weaknesses, and you can’t judge an OS based upon one or two devices that don’t work on it. Most of these guys are coding this stuff for free, especially in the BSD world (where there is much less commercial backing than Linux), so let’s not get all stupid about what our expectations are.
Maybe it’s a rough ideology, but I like the OpenBSD policy of “shutup and hack”…
Don’t like the hardware support? shutup and hack.
Want x_feature? shutup and hack
Don’t like how something is implemented? shutup and hack
But good grief, let’s end these petty squabblings.
Ive always been a fan of FreeBSD because of it’s elegant design, however ive used linux and windows as it’s easier for me to setup and support.
Although the big thing thats brought my interest back to FreeBSD is FreeNAS, the current version is great (apart from having only one share per drive) and having ZFS and other features in FreeBSD 7 is only going to make FreeNAS greater.
Can’t Wait!
nearly every os has a stable and a development ports tree. freebsd has not. this is the main reason i switched to openbsd. i really don’t care about zfs, dtrace and all other “cool” stuff. this is also the reason i won’t use freebsd in a production environment. keeping your system up to date means continuously fixing ports and thus angry customers.
freebsd itself has a stable tree, which is updated properly (only security patches and major bugs are fixed). the release cycle is also very good. why is this not extended to the ports tree as-well?
(yes i know there are release packages (no sec. patches, bugfixes etc), and please don’t tell me the ports tree is never broken)
Frankly, if all this stuff gets implemented, I suspect that FreeBSD will be in a position to take Linux on as the workstation/power-user-desktop of choice. PC-BSD has already proven that the FreeBSD kernel works well at the core of a desktop OS. With all this, PC-BSD will be ready to pwn…
I really hope that it won’t be PC-BSD…
they need to create ONE firefox..not a java-linux
and a java-linux-flash seperate version.It’s quite absurd.
Frankly I think that untill java source code is available Freebsd simply cannot compete.
DesktopBSD is in a much better position
“I really hope that it won’t be PC-BSD…
they need to create ONE firefox..not a java-linux
and a java-linux-flash seperate version.It’s quite absurd. ”
PC-BSD got both- Linux AND FreeBSD Java. BECAUSE in time of PC-BSD 1.0 there was NO FreeBSD Java available.
If freedom of choice is ABSURD then… ah, never mind…
> Frankly I think that untill java source code is available Freebsd simply cannot compete.
This saying is absolute nonsense – sorry. Inform yourself first.