[IP] more on READ more on Viruses

[David Farber <dave@farber.net>]

Begin forwarded message:

From: Johan <johan@ccs.neu.edu>
Date: May 24, 2005 1:20:41 AM EDT
To: dave@farber.net
Cc: Ip ip <ip@v2.listbox.com>
Subject: Re: [IP] READ more on Viruses


Christian Huitema <huitema@windows.microsoft.com> writes:


> The "small population" argument assumes that one can predict the
> psychology of malware writers. Incidents like the Witty worm show the
> limits of such predictions. In fact, one could just as easily make the
> opposite argument, "strength in numbers". Large populations are a  
> larger
> attack target, but they are also actively testing and developing
> defenses, and thus less likely to be swiped out by a catastrophic  
> event.

Well,

I dunno whether it's ease of infection alone, or a target-rich  
enviroment, that makes or breaks a virus.

A successful virus will be the one which as the highest chance of re- 
infection, which I'm going to posit is something like the product of  
the probabilities of finding an a suitable host and then infecting  
it, and for how long it can keep trying.

Windows viruses have a very easy time finding new hosts by just  
random guessing, while having a (hopefully) smaller probability of  
actually infecting the target, as it is likely running some form of  
virus protection.

Linux viruses (or Mac or OpenBSD) will in general have a harder time  
finding hosts at random, but may (?) have an easier time exploiting  
any holes found.

However, low population doesn't mean that it's hard to find a target.
For example, if I had an exploit for apache web servers, I'd have no  
shortage of targets.  I'm no firewall expert, but I wonder whether  
that wasn't the case with ISS. Firewalls are easy to find: just send  
traffic at a domain, and the firewall will intercept it.

The interesting part is that we've seen a marked shift in how viruses  
propagate.  Think back to the days of sneakernet and floppies; A  
succesful one had to be subtle - lay low and be stealthy for a some  
time before activating, else it ran the risk of not having propagated  
before detection.

In contrast, todays email borne internet viruses are a bit "blunt".   
I'd like to posit that the virus that will eventually sweep through  
the mac or linux communities will be more like sneakernet viruses  
than internet viruses.  Slow and subtle.

Johan

PS: I purposefully left out zombie nets so to not muddy the waters,  
but of course there's nothing stopping a population from  
simultaneously having subtle and blunt infections.  You just notice  
the blunt ones first.


-------------------------------------
To manage your subscription, go to
 http://v2.listbox.com/member/?listname=ip

Archives at: http://www.interesting-people.org/archives/interesting-people/