[free-sklyarov] Distributing AEBPR, contact to Dmitry?
Hynek "0A4h" Hanke
hanke at volny.cz
Tue Jul 24 05:10:41 PDT 2001
Hi
First thing I'd like to say is: ``Thank you all !'' I'm sad
I couldn't participate at these rallies, because it
wouldn't make much sense here in Czech Republic
in Europe.
I am interested in freedom of speech and these things
as deeply as most of you are. I have developed a simple
system to distribute such things and I consulted it with
one lawyer. It seems working.
The idea is simple. We will take a source data
(maybe binary of AEBPR). Make a package1 (p1)
full of random (good RNG needed) data so that
size(p1) == size (source)
Then we will make
p1[i] xor source[i] = p2[i]
In every step, we will randomly mix the i byte of p1 and p2
swap(p1[i], p2[i])
Now, it's mathematically impossible say if p1 or p2
contains random key or data. By law, it seems that key
cannot be judged as something dangerous. Key can
be used again and again on legal or illegal data.
Who knows... But is p1 or p2 key? That's impossible to
distinguish.
Then, using something like RAID, the information is divided
into 3 packages (p1,p2,p3), each one containing CRC
informations too.
Then we will put p1,p2 and p3 on diferent internet servers.
Now look, every package alone, p1,p2 or p3 is nothing but
random data (refering to the nature of XOR operation)
If someone forced me to delete p1 from server 1 (s1), because it
contains ilegal data, I can simply say him: ``Ok, I will delete
it so that you could look that it contains nothing interesting. The
information willremain avaible of course''.
(because p2 and p3 carries enought information)
I can distribute program (pr) along with these three packages
and say the user to link it (using this program) together.
Then I (as a distributor) am absolutely legal, because I
provide nothing but random data.
pr(p1,p2,p3) => source
The lawyer said to me it seemd OK. However, it doesn't
afect the responsibility of the end user linking the
3 packages together. But in this case, and maybe
in other DMCA cases, it seems the user made his program
out of random data doesn't have to wory about beeing
ilegal and beeing sued.
More information about the Free-sklyarov
mailing list