VMs: Constellation of ciphers...

[Nick Pelling <incoming@xxxxxxxxxxxxxxxxx>]

Hi everyone,

One quick thought on the whole constellation of different ciphers I'm 
thinking about.

A long time ago, I posted that I suspected that there might well be 12 
different constituent ciphers within the VMS' coding system: and I still 
suspect that this may well turn out to be true.

The point I was trying to make back then was that the system had resisted 
all assaults to date for a reason: and that the reason was probably that it 
incorporated multiple types of code within an overall framework - so unless 
you crack the framework, an assault on any one type of sub-code would still 
not produce practical results.

Here are a list of the sub-codes I think I can see now:-
(a) an abbreviating shorthand/tachygraphic code (probably vowel reduction)
(b) the space inversion transposition cipher (kill real spaces - insert 
fake spaces)
(c) a pair cipher, to hide the actual alphabet inside a fake alphabet
(d) a steganographic apothecary numbering shorthand (dain/daiin/daiiin)
(e) hidden index keys (typically between top line gallows, but could be 
anywhere)
(f) the <ch> generic shorthand completion token + disambiguation marks

If there are in fact 12 sub-codes as I suspected, then perhaps I'm at the 
halfway mark now (though I think labels will probably prove to be coded 
differently again). Perhaps one important thing to notice is that all these 
putative sub-codes appear (so far, anyway) to operate *in parallel*, not 
*in sequence*: that is, any given character is generally only coded with a 
single system.

I think this may be related to the overall robustness of the code: my 
suspicion is that this parallel coding system was an excellent compromise 
between error correction and security.

More as it happens... :-)

Cheers, .....Nick Pelling.....


______________________________________________________________________
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list