The
course of coding data is identical for all CODING programs:
§
The user gives the program an initial key entered by
the keyboard (direct key) or possibly by mouse
moves using a parameter dataset (random key) or
using a key dataset.
§
Possibly the user gives overlay
datasets and corresponding overlay parameters (overlay
information) as well.
§
Using these initial data, the program calculates a set
of parameters used to control the coding process and possibly stored in a parameter dataset.
§
If a parameter dataset is used, you have the
possibility to protect the usage of this parameter dataset against unauthorised
persons using a separate key (parameter dataset key).
Entering a key, a maximum of 256 or 512 characters using
CODING1 and 131,072 or 262,144 characters using CODING2 and 50,331,648 or
100,663,296 characters using CODING3 are shown and accepted potentially.
So it’s easy to see that for CODING2 and more
necessary for CODING3 you have to use key datasets
or overlay datasets in direct mode to get
approximately the entire range of keys being possible.
If a key covers more than 256 /131,072 /50,331,648 key
characters, the surplus characters are ignored. If on the other hand less than
this number of key characters are given, then the given key characters are
duplicated multiple and the result string is truncated to this number of
characters for further use. Nevertheless you should not choose to less and
homogeneous characters as key characters, because this, in extreme case (e.g.
only hexadecimal null characters), could be result in no encoding (!).
If you use a parameter dataset and before the
generation and final storing of the parameters for coding data, the quality of
the initial key or the derived key respectively is shown in form of statistics
(here: CODING1):
contents
of +--------- character-frequencies in
%: --------+ % binary zero
statistics: min.: max.: mean:
variance: ones: groups:
-------------------------------------------------------------------------------
Key: 0.00000000 1.56250000 (0.39062500)
0.38832041 49.36523437 93
For that
a key is much more identical to a “random key” the better it corresponds to the
following characteristics. That is
§
the number of null groups that is the number of
characters/elements not belonging to the key is close by the value 94 or 24,101 or 6,172,493
(CODING1/2/3) of all possible characters/elements,
§
the percentage value of binary ones is close by 50%,
§
the maximum character/element-frequency is close by
the mean value,
§
the variance is a small value.
Not having a rare extreme case, the results of
encoding, also using short keys, have no characteristics with regard to the key
length at all. Simply by searching the key a third person first will test a
shorter key probably, before it takes a longer input key into account – and
gives up then.
In principle all special characters excepting line end
characters (mostly generated by RETURN(¿) button or <carriage return><line feed>
or 0Dhex plus 0Ahex character) are interpreted as
components of a key of ASCII characters. During the key input entering
hexadecimal characters, line end characters may also be used to split key
information to more than one line. Line end characters in datasets comprising
an ASCII key are interpreted as components of the key too.