Actually, if y < 160, there's no guarantee that you'll get more than 1 file with the 
same CHK. Actually, there never is, but it becomes much less likely with y < 160, as 
some CHKs will have to be unused.
However, since Freenet rounds file sizes less than 8192 bits up to that in the CHK 
file-size field, it's not exactly relevant.

And actually, (2^N)/(2^160) is a good estimate for the number of N-bit files with a 
given CHK; getting larger than N is not a problem, getting larger than 2^N (the total 
number of different files with N bits) is the impossibility, but obviously will never 
happen.

-----Original Message-----
From: Lorrin Nelson <[EMAIL PROTECTED]>
Sent: Sep 10, 2003 9:47 PM
To: Discussion of development issues <[EMAIL PROTECTED]>
Subject: Re: [freenet-dev] CHK collision and denial of service

Yeah, I'm not sure this logic is getting anywhere. Lets say I want to 
block a particular y bit file (y < 160) with CHK x. If you start trying 
to find a file with CHK x by looking at all possible files, then of 
course you'll find the file in just 2^y tries, since you'll have 
stumbled across the actual file! But that's no use to use, since you 
need a _different_ file than the one you're trying to block. So you've 
gotta keep looking, and that'll probably take 2^159 tries as was said 
earlier.

I just made all this up so maybe I'm wrong, but it makes sense to me. :-)
_______________________________________________
Devl mailing list
[EMAIL PROTECTED]
http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/devl

Reply via email to