Review of OpenBSD 4.0

"I can't understand why anyone who has a choice would go with anything other than OpenBSD for a Web, FTP, email, directory, or NFS server."

I would certainly agree with most of that, though I have heard some grumblings about using OpenBSD as an NFS server.

i wrote an article called "New in OpenBSD 4.0" for a Czech unix/open-source internet portal Root.cz

http://www.root.cz/clanky/novinky-v-openbsd-4-0/ (in Czech)

just let you know, that Czech people love OpenBSD too :)

The only problem I got with OpenBSD 4.0 for the moment is that bsd.rd goes into ddb on my ibook. So I cannot install it on this computer and I have to keep 3.9 installed. Very strange..

OK everybody here loves to read nice things about OpenBSD. Nothing wrong with that.

How can this be called a "review"? I'd actually like to see a more objective piece. This was rushed out the door simply to get links from a site like this one to get more eyes on their ads. A real review would take longer to get out the door, and would back up some of its big assertions.

"In OpenBSD's case, the code is definitely high quality."

How would he know? Did he even look?

"Everything you get in the release is production-ready, secure by default (meaning the administrator does not have to lock down the system -- it is already locked down, and services must be individually enabled), and comes with possibly the finest integrated documentation in the Unix-clone world."

OK so he read the marketing speak. But if he had taken the time to read the finest integrated documentation in the Unix-clone world, he would see that there is a lot more to securing an OpenBSD system which is included in the documentation.

"As a result, OpenBSD's RAID and wireless network card support is exceptional"

How did he reach this determination?

"better than Linux's in some ways"

In what ways? Was this comparison part of the review?

"The only significant obstacle for desktop users is the lack of hardware 3D acceleration for video cards."

That's actually a pretty big obstacle for many desktop users. As much as he loves OpenBSD he should not be downplaying the importance of this point. Again, the lack of objectivity is quite apparent.

"and the GNU version had come to a point where it made more sense to rewrite it under the BSD license with security and portability in mind"

How did the author come to this determination? If he did not come to it on his own, he should be quoting an authoritative source. Again, this just looks like mindless regurgitation of what someone else told him.

I just got through half the "review" without any actual reviewing taking place. Why do we accept these fluff pieces? I reject it. This is not objective in any way.

The "Putting it to the test" section gives the first glimpse of anything useful. Talking about the added convenience of pkg_add since 3.9 was indeed very useful for someone who has either heard about or experienced for themselves the pain of upgrading an older version of OpenBSD.

The frank exposure of poor support for his new hardware was also refreshing after slogging through the first half.

"I've tried hard to find a significant weak point in this operating system, but there just isn't one."

D'oh. He just contradicted himself. After so much good progress he went right back into fluff piece mode.

"OpenBSD makes Unix fun and interesting."

Writers use words like "interesting" when they want to come up with something positive but hit a mental block.

"It's the only Unix-like operating system that you can build, customize, and update without running into strange problems, bugs, and growing pains."

The lack of objectivity is reaching unprecedented levels. Also, it directly contradicts the strange problems and growing pains that he has already run into within the very short scope of his "review".

"The other operating system I was testing while evaluating OpenBSD 4.0 was Fedora Core 6"

The only way he could have had a less relevant comparison is if he had picked Windows Vista instead of FC6. Something like Ubuntu 6.06 LTS (the "server" flavor) would have been a much better direct comparison.

"[...] I'm planning a migration strategy for my production server to go from Gentoo x86 [...]"

Oh sweet mother of Jesus... this guy is running Gentoo on a production server? Any hope for a shred of credibility just flew right on out the window.

Anybody that ever thinks it is appropriate to run an educational distribution like Gentoo on a production server should be drawn and quartered.

"The only complaint that I ever see people raise about OpenBSD is that it doesn't perform as well as Solaris or GNU/Linux when under heavy load. With high-performance computer hardware so inexpensively available these days, however, I can't understand why anyone who has a choice would go with anything other than OpenBSD for a Web, FTP, email, directory, or NFS server."

Wow he just contradicted himself in the space of two sentences. Either that or the "I can't understand" is simply an admission of complete ignorance of how the real world works.

If a reviewer brings up something like "It is commonly believed that OpenBSD doesn't hold up under a heavy workload as well as Solaris or Linux" (sorry, rms, no "GNU/" prefix from this objective reviewer) then dammit put together a test to present some illusion of objectivity. Make it actually look like you are reviewing the OS. Compare it in a few simple simulations of real load workload to other OS's.