[ale] OT: A funny one from Computerworld on XP SP2

Bjorn Dittmer-Roche bjorn at sccs.swarthmore.edu
Mon Aug 23 22:26:16 EDT 2004


On Mon, 23 Aug 2004, Jonathan Rickman wrote:

>
>> http://www.computerworld.com/softwaretopics/os/story/0,10801,9
>> 5390,00.html
>
> Thanks. Now if I could just figure out when buffer overflows became the most
> popular DoS attack, or when they became a DoS attack at all for that matter,
> I'd be all set. Any denial of service caused by a buffer overflow is either
> due to a poorly written exploit or is a mere side effect. If we surmise that
> a buffer overflow is by definition a DoS attack, then EVERY attack is a DoS
> attack and the water is now completely mud choked. The article is pretty
> misleading, and it's overall theme is factually inaccurate from the
> beginning. A buffer overflow is nothing more than a vehicle for attack. It
> is my understanding that the MS solution misrepresented here works by simply
> killing the attacking process, not shutting down the entire system. This may
> be an exercise in semantics, but details are important when discussing
> technical matters. Based on their misguided ideas of how things work, I
> wouldn't pay these two "consultants" to clean my windows (pun intended) for
> fear of them painting them instead.

I think you are absolutely right. My reading of it is that they are 
misusing the term buffer overflow. I think they are referring to 
overflowing some internal network buffer when an attacker sends a ton of 
data, so when the OS gets send more network data than it has buffers for 
it shuts down. I would be interested in a clarification on this one if 
anyone knows better. the link on the words "buffer overflow" give more or 
less correct definition, though.

 	bjorn

PS why am I suddenly receiving all ALE messages twice?



More information about the Ale mailing list