This article is a stub yet. It is just a placeholder for backlinks.

May be I will put some thoughs together about some UNIX caveats, like

• the (send)mail issue
• FHS
• package management overall
• the dependency hell
• market fragmentation
• duplicate program names (file collisions)
• partitioning and mounts (e.g. /var)

FHS is - sorry - bullshit. It isn't a careful deliberated standard, it is just a consolidation of the most used folders by UNIX' like systems in practice. Dot.

Useful folders in root:

/bin, /dev, /etc, /home, /lib, /mnt, /proc, /sbin, /usr, /var 

I know, there are many people out there who likes BSD's. To me it is impossible to understand their reasons, indeed IMHO they are a bit sadistic to themselves. Why using the easy (smart) way if a much harder, complicated way exists? Anyway, at the end it doesn't matter to me. So just some thoughts.

I want to exclude the available documentation. The official documentation is basically not bad, but it is just basics and if there some issues it is hard to find the solution. In other words: there is not enough (good) documentation.

Question: Are the BSD's are really ready for and arrived in the 21^th century? Or do they stuck in the Eighties of the last century?

The terror starts with the shell, by default  tcsh . Lets have a look at a weird example, which works under Linux (bash) well:

$> find /{usr/local,var}/*/bin/ -name qmail-smtpd -type f 2>/dev/zero | xargs dirname | xargs dirname | tee instdir.tmp | wc -l
Ambiguous output redirect.
$>

Great, an ambiguous output redirect because of the part  2>/dev/zero . Even weird that the arrow-up key re-formats the command - more exactly this part - to:  2 > /dev/zero .

Sure, for portable scripting the lowest common denominator should be used. But that doesn't solve the issue that similar commands have different parameters. Thus it is extra work to do required checks to find and use the correct options. Additional the readability is going worse. An example is  stat :

BSD:  $> stat -f %Lp         Linux:  $> stat -c %a 

Such things make the effort for OS checks for small projects which wants to be portable higher than the programming itself. Ok, it is not an issue of FreeBSD alone but all.

• IPv6 mapped IPv4 addresses are insecure: this is simply wrong and stated w/o any valid serious evidence
• update from one version to the next: this is at least an unacceptable effort (see thisdocumentation as an example - )