[plug] bringing insert-OS-here to its knees

[Leon Brooks]

Christian wrote:

> On Wed, Feb 28, 2001 at 10:48:44AM +0800, Leon Brooks wrote:
>> Windows 9X, definitely not. At last count, it was still possible to send 
>> semi-random ICMP to a 9X box and sooner or later it dies.

> Do you have any links to further info about this vulnerability?

Not to hand, but they do exist. Just like the M$ claim that NT syscalls 
were fully validated and so invulnerable to mistakes in syscall 
paramaters: somebody actually tried it and got through some miserably 
low average number of random syscalls before gutting NT each time.

> Microsoft has been pretty good about releasing patches in recent times
> so I'd be suprised if this problem exists and isn't fixed.  However,
> your words "sooner or later it dies" makes it sound like Windows normal
> behaviour and not really linked to any actual malicious attack... ;-)

I'd like to get all picky and point out the dichotomy between ``normal'' 
meaning ``how it is intended to be'' and either ``typical'' or 
``common'' meaning ``how it actually is.'' (-:

-- 
An employer once said, "What if I train my people and they leave?"
I say, "What if you don't train them... and they stay...?"