If you know so much about it, would you mind updating the Wikipedia article about ZFS that lists "Notable ZFS storage pool versions" with FreeBSD and Illumos both on 28.

I am currently using FreeBSD as my main OS, even on a recent (2010) laptop. It works great, for me. I have great control what is going on on my computer and I love the combination of a stable (as in "API/ABI stable" _and_ as in "upgrades do not break basic functionality") base system and very recent applications from ports.

Anyhow, I still have to agree with you that for most people, it is just not worse the time. For anyone else, I usually install Ubuntu. Ports are very powerful, but just not suitable for everyone.

If there were just stable ports (ports that come with a release and get only security fixes until the next release), one might come to the conclusion as the original article, but currently, you can either use release ports and live with the security holes (not a good idea to have an outdated browser and Flash plugin on the desktop), or you upgrade all ports very frequently.

PC-BSD might be a different story, I have not tried it in some time. Even though it brought me to FreeBSD, there were some good reasons not to use it anymore: How can you suggest users to use FreeBSD ports, if there is no PC-BSD PBI, and then wipe them at an upgrade? Probably not an issue anymore, but for me, that stuck.

bash is not preinstalled on many *nix systems. /bin/sh is ubiquitous and please, do not assume that it is bash. Fortunately, some Linux distributions undo that mistake. As for csh, it is widely regarded as being inadequate for scripting.

Powershell is about is ubiquitous as bash: It is present on most of the more widespread systems.

Looking for a suitable suggestion for a LaTeX editor and PDF viewer for Windows (cross platform would be a big plus) in our math department, I have tested several PDF viewers: Evince failed to render certain math symbols that did appear in Okular (and in Acroread and in TeXworks for what it is worse). Okular does not print. TeXworks lags some usability. Sumatra has not been tested, yet, but is next on our list -- your comment is not very encouraging in that regard. We have not found anything else that even advertises the functionality we need. (The build-in PDF viewer in Chromium maybe...)

Yes, print support in Okular would be great, especially since now that there are Kile binaries for Windows for the current version of KDE and Kile, Kile+Okular could be a nice cross-platform TeX environment. (TexmarkerX has severe bugs in the editor, TeXworks and Texmarker lags functionality in the editor, Emacs+AUCTeX is great but some people are simply scared by the Emacs shortcuts.)

People in our math department already use Firefox, Thunderbird, Matlab etc. that are all available cross platform. Some still do not even consider swapping from Windows to Linux for the change in the TeX environment. KDE for Windows could help getting people used to cross platform tools (but not without print support in Okular).

Solaris... there are alternatives. I wonder if ZFS will continue to be released to be used in FreeBSD.

OpenOffice.org... some project will build on it (and I do not need "Office"-Software, LaTeX does what I need).

Java... "Open" is not really done and the other license...

The only thing that I really worry about is VirtualBox. I have not found any other free Desktop virtualization that works.

Why do you think, "we FreeBSD-ers aren't getting Flash"?

I do have (the Linux version of) Flash 10 installed on my FreeBSD 8 amd64 systems and running it in a native FreeBSD amd64 Firefox. (Of course, it is usually blocked by noscript and flashblock.) A few years ago that might have been difficult to get running, but now it is just ports.

If we really want Flash is another story...

Probably you can do pre-7th grade math in one year, but you do not have much more time. With the beginning of puberty, many things suddenly become more interesting than learning new math.

I am working with selected -- so called gifted -- students of different age on math problems. I have given the same problem to 3rd grade and 7th grade students with the 7th grade students achieving not much more within 90 minutes than the 3rd grade students -- the problem did use knowledge from schools. The schools have failed in my opinion. Working on a different problem that involved some more rigorous proves (existence of Euler path'), the 7th grade students achieved more than the 3rd grade students on average (some exceptional 3rd grade student got most of it).

Either the article is right and the first six years of math education are more or less wasted even on the most skilled students -- or it is simply not the right approach that is used in school. As long as we do not teach "math" in school up to the high school level but only "computation", there are just cooking recipes, which tend to get boring, especially if the applications are flawed, too.

I have seen 4th grade students formulating proves by contradiction. Abstract thinking is possible in elementary school. I have seen many adults with university degree that fail on negating "C follows from (A or B)".

3rd grade students tend to be more open than 7th grade students, if you tell them that math without proves is no math at all -- because they have seen less so-called math.

The problems is that we do not teach math in elementary school at all!

Remote updates via the wlan are probably not very common. And I guess I could have put a brave "freebsd-update install && reboot" somewhere in the startup scripts that would have been replaced if that really succeeded. (Probably something a little more intelligent would be better.)

It appeared on the main ftp server on Monday and only an hour later on some of the mirrors. Now most of them got the bits. This is really not the time to stress the main ftp server more than necessary. The checksum files from the main server might be worse getting -- or better yet, wait for the official announcement that will contain them, too.

I did not read your blog using freebsd-update this time, but as far as I see, it would not have saved me needing hands on assistance for the system that I tried to update remotely with the last connection being a wlan. I added the appropriate lines to rc.conf before the update, but after the first reboot with the new kernel and old userland, the wlan did not come up. Thinking about it, nothing else could be expected...

On Sep-10, most debugging was disabled: http://lists.freebsd.org/pipermail/svn-src-all/2009-September/013399.html

On Sep-17, there was the last commit before 8.0-RC1: http://lists.freebsd.org/pipermail/svn-src-all/2009-September/013645.html

Anyone familiar with the FreeBSD development and release process would know that there are no fixed rules rules when certain stuff happens and there are no sweeping changes like turning off debugging between a late RC and the actual release. (Other debugging stuff like kernel and module symbols are kept for the release.)

According to some story circulating the net ( http://forum.ubuntuusers.de/topic/wo-kaufe-ich-ein-notebook-mit-linux-13-herste/2/ ), the Lenovo hotline in Germany denies that it is possible, but if you talk to a certain person at Lenovo, you will get a refund of 30 Euros for your Windows license.

I have not tried myself, maybe for my next laptop...

> I'm not sure how practical it is for any "programmer on the streets" to pay attention to this sort of thing.

These "any programmer on the street" guys hopefully never implement anything in the vicinity of crypto code.

You do not need to read the papers. Reading something like http://www.daemonology.net/blog/2009-06-11-cryptographic-right-answers.html -- if you happen to trust Colin Percival -- should be enough, if you do not try to be creative in what you use.

What is so bad about considering MD5 broken and make design choices because of that? Better than the other way around, if you are not in the field. How much more expensive is it to verify an MD5 and an SHA256 hash instead of MD5 only -- for many practical application, it is irrelevant. So why not do it?

/usr/local is for ports and packages. If you really need something outside ports to be available to all users, it might go there, too. Hopefully, you have a good idea of how to deal with updates. (Is there any reason not to make it a port?) /usr/bin is for the base -- the nonessential stuff that is not in /bin. No package should touch that since packages are much more likely to be broken than the base and should not interfere with it.

Sorry, your operation system has different ideas about hier(7) -- layout of file systems? Maybe that makes sense, too, but why do you think you know it better than the packages manager of another operating system / distribution?

Do these media for example include aljazeera? It is rather a selected media-driver viewpoint.

No, they cannot. Your statement does not get any more valid with this disclaimer.

The Islam accepts Judaism and Christianity as valid religions. There are many (fundamental) Christians that do not accept the Islam.

I do not think that the Islam is more tolerant than Christianity, which many Muslim claim. In most countries with an Islamic majority you should rather not even think about converting away from Islam. Anyhow, your claim that the Islam is "uncompromisingly intolerant" is rather close-minded (and very _certain_ media-driven).