Re: VMs: "Running code" -- I hope not...

[Elmar Vogt <elvogt@xxxxxxxxxxx>]

Maurizio M. Gavioli wrote:

> Well, if each character codes the "alphabetic distance" between two 
> characters IN THE PLAIN TEXT (rather than between the last coded 
> character and the next plain character to code), some structure would 
> remain; for instance, once coded, _rosa rosae rosae rosam_, etc... would 
> remain ostensibly a paradigm, as the 2nd, 3rd 4th letters of each word 
> would remain the same (if you "restart" at each word, even the first 
> letter would be the same).
>
> Wouldn't this match better the Vms. content?

Well, if the source was like that... unfortunately, by "running coding" you 
perform something like a differentiation of the text (in the mathematical 
sense), which leads to a shortening of characteristic passages -- ie 
characeristical triplets would be reduced to characteristic pairs, pairs to 
singles, etc.

So, for "standard" plaintext (for Italian, I used the Divine Comedy) which 
has only a limited amount of those sequences, they disappear completely from 
sight.

What we see in the VM is a _higher_ amount of characteristic passages than 
usual. That would point either to a very peculiar plain text (a rhyming 
dictionary?), or to a different coding scheme. ("Integrated encoding", 
rather than differentiating... how would that work?)

> On a different plan, I would expect repetitions of the same character to 
> be rendered in some special way (or even ignored) as, for any 
> (non-forgery) date which has been postulated for the Vms, 0 was not 
> reckoned as a valid number.
>
>         Maurizio

Again, unfortunately in none of the languages I checked (English, German, 
Italian, French) letter repetitions were outstandingly frequent, so a 
"Repeat" code isn't readily identifiable. Besides, you don't even have to 
introduce numerical thinking: If you abuse an Alberti Code Wheel with VM 
characters on the inner disk, the "Null" or "Repeat this" character is 
simply one of the letters on the disk. Alternatively, if you want to use 
numbers, you could use the number equal of letters of the alphabet on your 
disk (ie "Full Circle") for a "Repeat".

(I also found that you'd need a "synchronize" character which gives you an 
absolute position in the alphabet from which to continue, in case you miss 
letters on the vellum, or misread something. I'm not sure the gallows are it...)

Besides, the frequencies simply don't match.

With a running code, I get jump step frequencies between 3% and 6% for 
almost all possibilities, with most jumps having a frequency around 4.5% -- 
simply what you'd expect from a binomial distribution.

Compared to that, IIUC, the VM letter frequencies range as high as 13%, and 
have a long "tail" of rare, yet consistently observed letters. That appears 
to me to be a different characteristic.

As much as I would love to be the finder of the key, I'm afraid running 
coding simply ain't cutting it.

Tallyho,

	Elmar ("Integrated coding... hm...")

--
Elmar Vogt / Königswarterstr. 18 / 90762 Fürth / GERMANY
elvogt@xxxxxxxxxxx / Tel.: (++49/0)911 - 31 52 58
Agilmar von Sevelingen: VIS VISCERIS NON FERRE FERTUR (T.Doom)

"Patriotism is the virtue of the vicious." (Oscar Wilde, attrib.)

"I come from a country where being liberal is a virtue, and patriots are
viewed warily. YMMV."


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