Re: Randomness is the Tail - Not the Dog.



On Nov 4, 5:40 pm, amzoti <amz...@xxxxxxxxx> wrote:
On Nov 4, 3:38 am, austin.oby...@xxxxxxxxxxx wrote:





Randomness – a Transient Thing.

Random is a transient thing that is specific to one particular
circumstance called the experiment.  It lasts for the duration of this
experiment only and the data then reverts to being simply data except
for any other experiment it may pass later.  It is not in general, a
global property of a system like say data as prime numbers that are
unconditionally prime for all time.  The property of random is a
conditional one that is indeed true only by having its boxes ticked in
a particular experiment.  That of course may not be the sole object of
the experiment which will most likely have a quite different target.

Saying that data is random is quite wrong without specifying the
experiment that it relates to.  It’s like saying that something is
record-breaking without specifying the record.  The writer sees random
as a ‘lagging’ property in so far as it follows the experiment and can
only be ascribed to the data after the event.  The experiment
validates the data being random although the steps to make it random
must come from the user beforehand.

It is not possible to create random data (despite what the handbooks
say) and keep it handy for general use on the back of its being random
at all times when this is not yet proved.  I repeat it is subjective
to the eventual use.

There seems to be a mystique about the subject that suggests it will
take its place in mathematics some day and be fully vindicated by a
host of theorems and axioms that will enable proofs to be stated on a
par with modern number theory.  The writer hesitates to decry that
possibility but at the present time it does not seem likely it will
happen in scalar mathematics.

History has it that Pythagoras had one of his students executed by
drowning for postulating the existence of irrational numbers – may be
I’m next with regard to random theory and number theory becoming
partners.

The Boxes to be ticked:

1)The data is a string that has no rule – if it has rule it is a
sequence and cannot be random (despite what the handbooks say about
random sequences).  These are contradictory terms.

2) There must be equal probability between the elements of data with
regard to being correctly mapped by whatever the salient quest of the
experiment is.  Equal probability means zero certainty and an
adversary (assuming it is a hostile experiment) can only guess that
the outcome is any one of a vast possibility field that cannot be
verified even when the outcome is a de facto right one.

The ideal situation in cryptography is one in which some fortuitous
coincidence precludes all methodology that an adversary might use –
like the use of alphanumeric data as the data type for the cipher text
in the one-time pad type of cipher being prepared by the writer.

Subterfuge by scrambling while collecting data and then calling it
random like “ If I ask 100 people to verbalize the first integer that
comes into their heads “ or “the number of times a pidgeon pecks at
the ground “ or the “number of  splurges of radiation from a piece of
granite in a given time does not produce randomness.  These ploys are
useful in making sure that the collection of data is spontaneous,
impromptu and unrehearsed and that it will not go into an experiment
with some glaring baggage called visible bias (say) that could be
developed by an adversary into structure.  They do not on their own
provide random data as a result of such a scheme. This has confused a
lot of readers.
Dat + Structure = Information
No structure no give-away information

The foregoing under analysis.

To get to the point, scalars as numbers or scalars as letters is
totally unsuitable as data for encryption work at least, if not in
fact for any and all use of randomness as a devious ploy.  The earlier
foregoing remarks are intended as rhetoric more than accusation but it
is necessary in the writers view that it be generally realized that
the current cryptography that is based on numbers will never become
strong and a change of course is essential now and quickly.

The writer is developing the first implementation of an ASCII
modulated OTP cipher that is truly ground breaking as Theoretically
Unbreakable cryptography but this is only a lucky fallout of computer
science and a data structure called an 'enumeration type' in modern
programming languages.

 It is not mathematical to depend on something fortuitous like this
and indeed something better as data is to hand that could elevate
randomness into that state where it has respectable rights as theorems
and proof comparable to number theory.  These claims of course can
guarantee the cryptography that this data type then underpins as
ciphers.

It is necessary to “Up” the entire ball-game by an additional
dimension from number lines in two dimensions to three-dimensional
space.  The data of three-dimensional space is automatically random at
all times even for one-off elements that are stand alone operands and
not even in a string. The irony of this fact is that randomness is
superseded by the one-way function that is available in vector data
methods and may not even be required.  The whole of space is a random
number domain.

This is something that simply has got to come into vogue in time but
it needs a change of attitude all round by the establishment now.  It
requires an effort that may not be within the ken of pure
mathematicians.  A radical change is needed now even to hopping the
ball in a different court such as Applied Mathematics and Engineering.

Cryptographers will simply have to learn the methodology of Vector
Methods in applied mathematics and apply it to cryptography.  The
result will be revolutionary and indeed raise randomness in
cryptography to a very acceptable branch of applied mathematics that
has its very own laws and proofs. It is waiting to be done – Adacrypt..

PS Think about it – any integer of the infinite set that exists may be
assigned to any one of an infinite set of points in space instead of
sitting on a single number line that has arbitrary direction as it
does in current cryptographer.  There is simply no argument against a
change to this vastly greater domain of data that has intrinsic
randomness due to being a set of non-parallel vectors.http://www.adacrypt.com

General Vector cryptography and in particular “A New Approach to
Cryptography”.

Have you read this paper and used this code to test anything you have
done?

http://www.iro.umontreal.ca/~simardr/testu01/tu01.html

It includes results for some crypto rng's.

Enjoy ~W- Hide quoted text -

- Show quoted text -

Have you read this paper and used this code to test anything you have
done?
I wouldn't touch rngs of any origin with a barge pole.

The Vector Ciphers in my portfolio are mathematical provable and have
been tested only on messages to myself - They are on offer for
external beta testing by anyone - about six of them.

I am working on a " ASCII Modulated OTP" at present and that will be
up for beta testing also fairly soon - a few weeks.

If any one is interested in either of these then let know and I will
forward the necessaries.

I will read that article on the link later - Many Thanks - Austin O'
Byrne
.