Re: How to destroy windows 2000 :)

From: Shannon Jacobs (shanen@my-deja.com)
Date: 12/02/02 

From: "Shannon Jacobs" <shanen@my-deja.com>
Date: Mon, 2 Dec 2002 22:50:50 +0900

First, your "security problem", even allowing for your exceedingly poorly
thought out description of it, is solved. It is an old problem, and
intelligent people have analyzed and solved it, and provided proofs of the
validity of those solutions. I am not smart enough to have written those
proofs, but I am smart enough to find them, mostly understand them, and I
probably could even implement them, though I'm not programming in those
areas in these years.

Second, your proposed solution is provably wrong. If you have already made a
clumsy mistake that has compromised the operation of your program, then any
security "precaution" or "countermeasure" you propose after that point is
already known to be toast. Once a system has been compromised, there is no
way to prove it has been repaired or that it is intact in any particular
portion. That follows fairly directly from Godel's Theorem for any system of
sufficient complexity. (Okay, I'm glossing over one step. Though many
computer languages are over the threshold, I have not seen a complexity
proof for the Windows API as such. I strongly suspect it has far more than
the required degree of complexity, but that should be checked, because in
some ways the expressiveness of Windows is quite constrained.)

"MacroMad" <username@email.co.yu> wrote in message
news:ased0l$qm1he$1@ID-131597.news.dfncis.de...
<snip>
> 1. I'm not from Germany

Yet another good example of your stupidity. I didn't say that, and in fact
there is evidence you cannot be German. However, Germany is where the logs
to identify you are located.

<snip>

> 4. I don't like you

Thank you. I'm quite pleased to be disliked by such people as yourself.

> 5. YOU ARE A ***

Leave my girlfriend out of this.