complacent wrote:I mean, come on people!! This is what, like the
[/b]SECOND[/b] security patch released in the past
10 years?!?!
You know, I've actually been lecturing on this all week. I've been really surprised at how much faith people put in "secure systems." OpenBSD is phenomenally secure. They have more or less the right idea for how to create a secure system (DISA tends to recommend that you "turn off things you don't need." The correct approach for securing systems is to only turn
on things you need, leaving everything off. OpenBSD calls this 'secure by default', and it is the right approach).
However, even today (this is the last day of the class, they're getting the exam tomorrow), I had to give like a half hour lecture on why the very secure isn't. They looked at me in amazement when I explained:
"So let's imagine the world's most secure system. You have a network and applications that are absolutely
impenetrable. You're storing all your fancy mission data on it, knowing that your targeting data and intelligence data are absolutely safe on the system.
At the end of the day, you leave the office, comforted by the fact your data is safe.
When you get home, you say hello to the wife and kids, and fire up your email client and answer an e-mail from your mother asking how you're doing and what you've been up to, by voicing your frustrations with the way the war is going, including mission data."
I thought this kind of thing was glaringly obvious.
So, sure, OpenBSD is secure. But, human elements make it only as secure as the least secure part of it. Pricks like Theo screaming from the hilltops about how secure it is (making it a huge target; I'm sure we remember the *GOBBLES* days – these were aimed squarely at Theo, not just the OS), and idiot sysadmins who install it thinking that by running the world's most secure operating system, there is
no way their data can be compromised.
See also: SELinux. Trusted Solaris.