Re: 2 forwarded messages...DNSEXT discussion of Day and Kaminsky

On Fri, Feb 20, 2009 at 05:24:24PM -0500, Dean Anderson wrote:

> > > You collaborate with a known blackhat, Kaminsky. That makes you a
> > > blackhat, too.  (Media3 v. MAPS) That is a fact. 
> > 
> > Am I blackhat, too, then, since I collaborated with Kevin by making the
> > patches[1]? I can't wait to reap the ill-gotten rewards of my life of
> > crime.
> 
> I think that all depends on what you do, I think.  I can see that it

I think you missed my point: if collaboration with Kaminsky makes Day a
blackhat, then wouldn't collaboration with Day make me a blackhat? In
other words, how can I (or anyone) possibly "launder the blackhat
taint", if doing so makes me a blackhat?

> > So do I still get to launder the blackhat taint away?
> No. You don't.

OK, well at least that is logically consistent.

> I have a legitimate analysis of the problem; you have seen it;  but you
> STILL (if above is any indication) can't even repeat it back to me
> accurately.  Until you can show that you understand my analysis, or some
> analysis, I can't really have rational discussion with you.

Well, I certainly agree that it doesn't seem like we're having rational
discussion; I am having a very hard time understanding the attack you
claim exists. If it were just me, I might think I had a reading
comprehension problem. But given the other responses, maybe you are not
being clear enough?

Paul was nice enough to post an attack scenario based on some of your
comments; I have just replied with my own analysis which, unless there
is an error, shows that it is not a problem.

> Until you can accept that character matters, I can't trust you.

I'm not asking you to trust me. I am asking you to look at the patches I
wrote and decide whether you want to trust _them_ or not[1]. Patches do not
have character.

-Peff

[1]: Actually, I don't even really care whether you trust the patches,
or whether you install them on your system, or whether you recommend
that other people install them. What I do care about is whether there is
a practical attack that works better against qmerge-patched dnscache
than it does against stock dnscache. You seemed to indicate that you
know of one, which is interesting to me (and presumably to others on the
list).