This Improvisation Will Buy Time Alright But….



Many thanks to every body who has subscribed to my new understanding
of what is going on in the real world of cryptology.

I think I’ve got it right now and in truth it is what I suspected all
along but wasn’t sure. It seems to me now that cryptographers are
using up the almost ‘spare, less-used capacity’ of ASCII in the form
of the control character byte values to encrypt industrial data
(images and sonic) that is not describable as lingual or language data
under normal ASCII terms of reference but is necessary to encrypt just
the same. The effect of this on ASCII seems to be no more than
academic to hands-on people in the industry.

To do this they use the convenience of modern computer architecture
being in bytes and therefore is already keyable directly or indirectly
to represent words of other non-lingual data as a convenience. So
this is what you might call a convenient misuse of ASCII but one that
works.

Taking the broad helicopter view it is an adaptation of ASCII that
goes outside of the remit of ASCII but one that is favoured by all
hands. Don’t forget that the operative word is information
interchange with the accent on verbalised human intelligence in
ASCII.

If this broad analysis is true then that’s ok with me but I honestly
think it is only buying time and will, just like AES itself, be always
computer-dependent and therefore prone to brute force. It cannot stay
the testing course ahead in terms of viability.

Design orientated reader mindsets should see it as complexity-
theoretic cryptography that will only last as long as computer power
doesn’t get large enough to brute force it.

What this is really saying is that ASCII needs to be either extended
or rewritten completely and the standard keyboard is now inadequate to
cover all modern cryptography but before going down that extreme road
of suggesting change, some other things need to be considered.

In my view block ciphers work by confusion through diffusion of
strings of binary digits and they are only complexity-theoretic
instead of the more versatile number-theoretic. These ciphers are
breakable by computer power whenever that becomes available at
sufficient levels. Adopting byte representation as alternative binary
base number systems that are just as transparent as the universal
number system in base 10 is no better for encryption transformations.
The name of the game is still security and using byte representaion
that happens to be conveniently to hand is not a lasting solution.

It is the sheer transparency of the universal number system that is
the weakness of current cryptography. Encryption in bytes is merely
substituting the universal number line for one that is conveniently
readymade and compatible with modern computers but it is doing nothing
to alleviate the transparency of the numbers it uses albeit in bytes
now instead of denary base.

All that this is going to do is perhaps buy a little time but the long
term solution in my view is still in using a plethora of different
number-lines as I do in denary base as the norm in cryptography.
These latter are impossible to cryptanalyse.

I believe this simply has to come to pass if cryptology is to be
immune to increasing computer power in the future.

From a management perspective I reckon that what everybody is saying
in the threads without realising it maybe is that ASCII isn’t
sufficient any more and something else is needed soon. True or False?

Would Uncle Sam agree to that I wonder.

Again, many, many thanks for your much valued help.

- adacrypt
.



Relevant Pages