Certain bases are especially good options in which to generate native
element streams; some bases are poor choices The idea of going
directly to character sets to be used in stream encrypting of
plaintexts in the same set is an old one. Base 26 was a good choice
for classic code clerks used in a classic algorithm but we are hardly
restricted to that.
My research has demonstrated to me that bases can rated as to
mathematical usefulness, whereas rather long cycles of such characters
can be efficiently generated. Beyond a simple progressively
traditional indexed set, the elements can easily be scrambled to an
obscure sequence for running key and reindexing the plain characters
as well.
Working in novel number systems with a custom javascript pRNG using
simple essential common statements that sidesteps the lost of or
compromising of any standard generator which might be withdrawn or
changed on governmental whim. I do favor in my implementations open
length keys and seeds, build-in tests to see that tens of thousands of
characters are easily generated in a key loop, and multilevels of
processing are usually allowed even as codegroups can be an optional
style when formats.
Javascript allows immediate field customization of such programs by
the qualified whereas default sequences and keys can be easily
edited. I make no claim that any examples should not be so changed,
public domain,...enjoy. Remember that the nature of my examples
explore a wide range of strength so that cryptalysis might be
facilitated in some of them on purpose.
The discussion here, if any, might best initially be directed to the
effects of reordering of the character set and the several levels of
complication that might be used. The tests of many bases indicate that
binary derivative bases are usually very bad choices as closed loops
in them often prove trivial. I'll tend to keep any ct examples as
brief as is reasonable as I understand the direction here is not to
abuse bandwidth.
|
