Re: 1.7: Problem with Vista64b ACLs and sockets

[Warren Young <warren at etr-usa dot com>]

Dave Korn wrote:
> FUDmonger Gibson 

He does go to extremes sometimes, but that's his (self-appointed) job. 
In any sort of advocacy, it takes extremists on both sides to help the 
rest of us find the middle.

The main criticism I have of Steve Gibson is that he frequently forgets 
that security is a people problem, not a technical problem.  The 
software has to do the right thing, of course, but ultimately, if people 
want to roach their systems through negligence, no technology is going 
to help much.  Tricking ignorant users into running malware has to be 
either the #1 or #2 way worms get on PCs.  (It's a toss-up between that 
and all the remote code execution and privilege escalation holes.)

> We now have the benefit of hindsight, and it's made exactly _how much_
> difference to the usability of XP machines as botnet drones sending spoofed
> packets in DDoS attacks?

Err...disallowing raw socket access to all users doesn't fix the people 
problems and the remote root exploits, so it's a bust?

How about, instead, we educate the users and arm-twist Microsoft to fix 
all those holes so that it actually matters that raw sockets are 
restricted?  If more people listened to Security Now, there'd be a lot 
fewer bots.  I'm not saying that people should follow 100% of Steve's 
advice.  Just getting cluebies to stop clicking on links in spam and 
"NAV2009" popups would help loads.

Don't forget, what Microsoft did here is finally follow the standard 
behavior on Unix-like systems, which we're all supposed to really like 
here, right?  /bin/ping on Linux is setuid, no doubt for this very 
reason.  Does Windows not have something like setuid?  If not, there's 
another legitimate reason to criticize Microsoft.