RE: Steganography

From: Eugene Leitl (Eugene.Leitl@lrz.uni-muenchen.de)
Date: Fri Sep 28 2001 - 00:56:04 MDT


On Thu, 27 Sep 2001, Harvey Newstrom wrote:

> I don't know about this specific case, but this scenario has been
> repeated so many times I don't know why people keep trying.

I don't know either; it's completely pointless.

> Brute-force has been used to crack "uncrackable" codes by combining
> thousands of Internet computers over and over again. The
> distributed.net network is running close to 200Gigakeys per second!

This is getting tedious. You, of course, realize that these challenges are
chosen so that the current hardware base is sufficient to brute-force the
key space in a few months? That each bit added to the keysize doubles the
search space? That the key size alone doesn't say a lot, unless you also
specify the computational requirements of the algorithm.

> Nothing is uncrackable these days, even with brute-force. Encryption

Sorry, but this is compleat bullshit. You know how the crunch of the
average computer grows, how many of them are around, and can choose the
key size so that encryption is still sufficient speedy whereas decryption
by brute force will remain prohibitive for n years. If you drop the
"sufficiently speedy" requirement, n can become rather large.

> is a temporary solution at best. We must assume that at some date in

No one is talking about encryption for an eternity, though one can come
close as soon as we're in the stage of omega hardware, where you can only
scale up by scaling up volume.

Secrets grew really stale really quick. In financial crypto a few minutes
will do, in the military it's been typically decades, but things move so
fast now the secrecy window is shrinking.

> the future supercomputers or networks of computers will scan all
> Dejanews/Google archives and decrypt all stored messages in the
> history of the Internet. No one should send anything encrypted over
> the Internet that you don't want read publicly in the future.
>
> RC5 48bit Challenge, February 1997: (cracked in 13 days)
> <http://gilchrist.ca/jeff/distrib-rc548.html>

RC5 [Riv95] is a fast block cipher designed by Ronald Rivest for RSA Data
Security (now RSA Security) in 1994. It is a parameterized algorithm with
a variable block size, a variable key size, and a variable number of
rounds. Allowable choices for the block size are 32 bits (for
experimentation and evaluation purposes only), 64 bits (for use a drop-in
replacement for DES), and 128 bits. The number of rounds can range from 0
to 255, while the key can range from 0 bits to 2040 bits in size. Such
built-in variability provides flexibility at all levels of security and
efficiency.

Do you see the difference between 48 bit and 128 bit? Not to speak of 2
kBit?

-- Eugen* Leitl leitl
______________________________________________________________
ICBMTO: N48 04'14.8'' E11 36'41.2'' http://www.lrz.de/~ui22204
57F9CFD3: ED90 0433 EB74 E4A9 537F CFF5 86E7 629B 57F9 CFD3



This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 08:11:02 MST