FreeBSD 6.2 review

[Morbius]

After installation, FreeBSD is left with no real make.conf or rc.conf

This is silly. The most sensible defaults for both of these is an empty file. rc.conf is only an overide for /etc/defaults/rc.conf. All of the default software is built with an empty make.conf file. As far as possible FreeBSD prefers to setup its config files as overrides.

Why is FreeBSD using a shell that doesn't at all compare feature-for-feature with more modern alternatives like zsh or Bash, and was designed to be used on terminals that could reasonably be found in museums?

It's there for running Posix shell scripts Why on earth would anyone use the default shell when it only takes a few seconds to change it? It's like complaining that your car's radio was preset to the wrong station. It's not a serious issue unless you are a complete UNIX beginner, and if you are, FreeBSD isn't the place to start.

Less hassle for the JDK and JRE

There's a pre-compiled version than can be used if you want. But the problem with source code is purely down to Sun.

[infofarmer]

Jem and others,

maybe it's not clear in the heat of discussion, but none of us are on the opposite sides of whatever. We all use FreeBSD more or less often, most of us like it, and some of us even love it. Jem's review was a bit straightforward, but believe it or not - it can be more helpful than sheer praise.

We might want to see every issue become a problem report in our gnats database, but there's nothing wrong in writing "FreeBSD sucks" articles. When I read such a review, I don't see an attack, I see objectives. This user is not completely satisfied with our product - it means there's room for improvement.

Let me take a chance and thank Jem and others on behalf of the FreeBSD developer community for the review and the discussion it inspired. Jem, it would be great if you took a more formal, traditional way of participation in our user community, but at any rate I would love to see more of your articles.

[Valour]

Originally Posted by infofarmer

Jem and others,

maybe it's not clear in the heat of discussion, but none of us are on the opposite sides of whatever. We all use FreeBSD more or less often, most of us like it, and some of us even love it. Jem's review was a bit straightforward, but believe it or not - it can be more helpful than sheer praise.

We might want to see every issue become a problem report in our gnats database, but there's nothing wrong in writing "FreeBSD sucks" articles. When I read such a review, I don't see an attack, I see objectives. This user is not completely satisfied with our product - it means there's room for improvement.

Let me take a chance and thank Jem and others on behalf of the FreeBSD developer community for the review and the discussion it inspired. Jem, it would be great if you took a more formal, traditional way of participation in our user community, but at any rate I would love to see more of your articles.

I didn't even see it as a particularly negative review. Heck, I really like FreeBSD, which is why I continue to write about it. While it is a decent operating system, it does need a lot of small improvements to make it more competitive. What kind of a reviewer would I be if I didn't point out the problems I had while testing the operating system?

One thing that a lot of people who have commented on this review don't realize is, I've written about every FreeBSD release since 5.1. Several of the reviews have made Slashdot at various times, but somehow it seems that a lot more people are suddenly aware of this one over the others. I also wrote benchmarking articles comparing the scheduler performance of ULE vs. 4BSD, and i386 vs. AMD64.

I am really glad to hear from you, and that my review was taken seriously by the FreeBSD development team. I used to be on the AMD64 mailing list back when the AMD64 port was still having a lot of growing pains, and did file some bug reports. I wish I could help with other problems, but I'm not really familiar with operating system programming. If you can think of a way that I can help, let me know. Perhaps I can donate the Microsoft keyboard that didn't work properly in FreeBSD to a programmer who can fix the problem.

I am working on a guide for O'Reilly called The FreeBSD 6.2 Crash Course. I've already committed to donating 10% of my royalties from it to the FreeBSD Foundation. It probably won't end up being much, but it's something.

Anyway, thanks for your note.

[Oopla]

It always seems to be the "It's good, but.." reviews that draw out the flamers. Reminds me of the Dragonfly BSD review response.

[Valour]

Originally Posted by Oopla

It always seems to be the "It's good, but.." reviews that draw out the flamers. Reminds me of the Dragonfly BSD review response.

I'm fairly convinced that only psychos use DragonFlyBSD. That OS is falling into a niche for clustering, which is fine, but nothing I'm interested in writing about.

There is nothing in the computer world that is not "good, but in need of improvement" from at least one person's frame of reference.

[Valour]

FreeBSD's advantage is that it is a cohesive, singly-developed system. It is not "GNU and Linux" and is not open to meddling by distributors. So in theory it should be a more stable and efficient system. Unfortunately, in much of its recent history it has been nothing less than a disaster. It only began to find redemption in the FreeBSD-6 series.

The big problem in FreeBSD's past was that the developers planned huge changes to the kernel that did not work out as intended. This is anathema to the reason why the Linux kernel has been so successful -- it proceeds slowly and with short, easy goals in mind. The same can be said of OpenBSD -- realistic goals that are easily met to show long-term progress.

I see that FreeBSD-7 has a number of huge changes in store, and that worries me. I would like to get away from Linux and back into FreeBSD on my workstation, but I can't do that if every other release is going to crash every hour due to multi-threading problems, or hang for long periods of time because of disk I/O code problems, or drop network connectivity because the network driver crashed. I also need hardware 3D acceleration, which was not possible for a long time because the Nvidia driver for FreeBSD was old, due in part to some structural limitations in the FreeBSD kernel.

So things are pretty good with FreeBSD at this particular moment in time. But let's see what happens with the next release.

[Pigmagropyirm]

There is highly interested business proposal!
After years of research and development,
Planet Resource Recovery, Inc.
showing for the public their nanotechnology project
enrolled
under the trade name PetroLuxus.
This unique
Green technology is biodegradable,
nontoxic and nonflammable, and its molecules are so microscopic that it can penetrate virtually any natural or manmade material, and break the bond between those materials and hydrocarbons. PetroLuxus is being utilized to increase production 100 to 500% in marginal oil wells, and is also a silver bullet for property owners, as well as state and federal environmental agencies in the remediation of contaminated soil and water, making it possible for the first time to recover hydrocarbons and precious metals from soil that would otherwise be hauled away and burned to ash.

In case of any questions, see http://www.planetresource.net