Submitted by The NetBSD team has released the second release candidate of NetBSD 4.0. “There have been many fixes since the previous release candidate, RC1. The most important ones are: ICH9 support in wm(4); enhanced Speedstep support for VIA C7/Eden and amd64; many bugfixes for IPF; FAST_IPSEC fixes; wpi(4) bugfix; proplib local DoS fix; and much more.”
Since the (in)famous benchamrking between various Linux and BSD kernerls http://bulk.fefe.de/scalability/ a few years back there has been continuous work on NetBSD.
See the following for a flavour of things to come in SMP performance:
http://www.feyrer.de/NetBSD/bx/blosxom.cgi/nb_20071001_0149.html
Take a look at this:
http://mail-index.netbsd.org/tech-kern/2007/10/05/0000.html
🙂
Fefe is a self-declared Linux fanboy, without any real knowledge of *BSD.
The mentioned benchmark is wrong, because they did not deactivate some debug functionality in FreeBSD current:
http://marc.info/?l=netbsd-tech-kern&m=119142054028425&w=2
This is the correct one.
Hmm… resorting to namecalling, eh? Very mature… suuure. Stop being so infantile and elitist. It’s exactly that kind of behavior that keeps *BSD as the could-have-beens despite the technologically superiority on most issues.
EDIT: Calling him names is quite unfair because he is praising the *BSD (especially FreeBSD) to the skies, and the few valid complaints were fixed quickly.
Luckily the devs in the respective *BSD’s (especially NetBSD and OpenBSD) treated Felix much better than the *BSD-zealots like you.
If you are talking about Felix’ benchmark you are incorrect. Debug functionality was disabled in his benchmark 4 years ago. The results were btw. validated by FreeBSD, NetBSD and OpenBSD developers. Only infantile elitist fanatics like you discredit his work, which btw. has been confirmed by the later work you also refer to.
Edited 2007-10-05 19:38 UTC
>Hmm… resorting to namecalling, eh? Very mature… suuure.
Did you ever read the benchmark? Probably not.
>Calling him names is quite unfair
Mentioning this crap at all is nonsense. Even most Linux fellowers call him names, but they aren’t as polite 🙂
>If you are talking about Felix’ benchmark you are incorrect.
No I don’t. You should actually read something first, before lamenting.
>which btw. has been confirmed by the later work you also refer to.
Learn to read mature boy.
This,
http://marc.info/?l=netbsd-tech-kern&m=119142054028425&w=2
has nothing to do with your enfant terrible. Search the mailinglists for fefe (Felix von Leitner) in terms of Debian and so on. This Linux (Gentoo) Fanboy axes *any* different opinions with harsh words.
http://rrr.thetruth.de/2007/08/writing-a-bug-report/
Read this, it’s about fefe. It’s an immature boy, get it.
http://www.monkey.org/openbsd/archive/misc/0310/msg00840.html
Read some other opinion about this so-called test aka nonsense of fefe.
>The results were btw. validated by FreeBSD, NetBSD and OpenBSD developers.
Do you need more proof for you nonsense-saying? I guess not.
“OpenBSD 3.4 was a real stinker in these tests. The installation routine sucks,”[sic!]
–fefe
As I said immature behaviour par excellence.
Have a nice day and please do your homework first next time before flaming mature people. Thx 🙂
Edited 2007-10-05 20:25
I tried NetBSD 4.0 RC1 last week and ran into some problems and gave up:
1. Is Xorg binaries availale with NetBSD yet? Everyoen else seem to have on but NetBSD seems to be stuck in XFree86. Some of the newer Intel chipsets don’t work with XFree86.
2. I had serious problems (and I was not the only one based on Google search) with the USB mount of Flash/thumb drives (this was the reason I gave up). I was able to mount the MSDOS formated flash drive on every other OS but not NetBSD? It kept complaining about Invalid Parameter. What else other than “mount_msdos /dev/sd0e /mnt”. (Yes there was a DOS partition on e: based on disklabel output)
3. NetBSD would not recognize DOS partitions on some other drive even though every other OS did. Disklabel did not even show the DOS partition even tough Linux and FreeBSD mounted it just fine.
I really want to run NetBSD but having used FreeBSD and Debian, it was a lot of work to get the same job done. I am going to give RC2 a try and hope some of this is fixed. (Is it?)
Anyone else with such issues?
From what little I have used NetBSD, its is indeed very quick and shows great promise. It is also very easy to maintain (hey, they got Xen working on it and the FreeBSD guys have been at it for more than 2 years and don’t seem to get dom0 working properly).
Another download underway….
-D
>hey, they got Xen working on it and the FreeBSD guys have been at it for more than 2 years and don’t seem to get dom0 working properly
Of course, because they did not work on it.
1/ modular-xorg is available in pkgsrc. You will find binary package from pkgsrc2007-Q3 when NetBSD 4.0 will be out.
You can still compile it using pkgsrc or use an odler binary.
2/ I never saw this issue. You may ask the question on a netbsd mailing-list and possibly report this issue (if this really a technical issue).
3/ You may use mrblabel to add “automaticly” the different partition reading the mbr. After that, you will find the different partition in your disklabel and you can mount them.
Bests,
1. Get it through pkgsrc
2. This was probably due to a 2kb-sector FAT partition, which we now handle fine (e.g., iPod Nano).
3. Description of the issue is a bit vague, you might want to ask a list with the fdisk output of another OS and the output of fdisk and disklabel on NetBSD.
Quentin Garnier.
Thanks. I think you might be right. I will investigate further.
Another question: I have a laptop with 2 cores (intel 7200 cor 2 duo). Which kernel do I run? GNERIC, GENERIC.MP or GENERIC-LAPTOP?
When I run GENERIC, it says I have 2 processors and that the second one is application processor. cpu0 is started but cpu1 is not (from dmesg).
Thanks.
-D
You run the GENERIC.MP ( Multi-Processor)
As project_2501 said above, there are huge improvements going on in NetBSD, specially in SMP, wireless drivers (following OpenBSD steps), virtualization, etc. OpenBSD 4.2 is coming too.
There are a lot of hype caused by many operating systems (mainly Linux distributions) in OSNews, ranging from minor visual changes in Ubuntu (Linux Mint?), forks that in fact are the same as the system forked, etc.
This one is different. This is real computing. Congratulations.
“””
“””
If they are making rapid progress, that is good. But in all fairness, NetBSD desperately *needed* a huge amount of progress to regain relevance in today’s OS landscape.
See this candid post from a NetBSD founder, written about a year ago:
http://tinyurl.com/q2366
Edited 2007-10-07 14:57
I read this article. But the fact (this benchmark shows it) is they are really making huge improvements whatever is happening in “The Foundation”.
NetBSD is one of the most performance-addict OS in every aspect, even with few developers and even being not enterprise supported like Linux. What make an OS relevant? What is not fair is call an OS like this as irrelevant just because a small “footprint” of users.
You can follow up on http://mail-index.netbsd.org/tech-kern/2007/10/
For those who are taking this personal take a look at FreeBSD here:
http://mail-index.netbsd.org/tech-kern/2007/10/05/0019.html
and here
http://mail-index.netbsd.org/tech-kern/2007/10/05/0025.html
Gotta love the last graph tho.
>For those who are taking this personal
http://bulk.fefe.de/scalability/
Nobody cares anymore about this ‘test’ of this Linux-guy.
Those later benchmarks are just for testing purposes in terms of MySQL (and not a self-build wannabe-server) and scalability.
http://marc.info/?l=netbsd-tech-kern&m=119142054028425&w=2
This is the latest benchmark (after the disabled debug in FBSD current) and as you see, it’s a bench between *BSD, nobody really cares about Linux.