Re: D!

[Mark Perakh <perakh@xxxxxxxxxxx>]

There are several transliteration schemes converting VMS alphabet into other
characters, for example Currier's scheme transliterates VMS into Roman
characters plus 9 numerals plus asterisk. All those schemes are, of course,
arbitrary.  Now, the number of possible schemes, while enormously large, is
finite. What if to write a program that will arbitrarily assign  Roman letters
and numerals to VMS characters, and then apply the alphabet obtained that way to
some relatively short segment of VMS text, then change the assignments, and so
on and so on, each time trying to convert a segment of VMS text into a different
set of the same Roman characters, and in each next trial altering the
assignments of letters.  It may take billions of trials, but we have all the
time in the world, and possibly with the assignment number zillion some readable
text in some recognizable language will show up. Is such approach feasible? Best
to all.  Mark

John Grove wrote:

> Don't ask why, but as I was doing some character replacements I found a
> rather peculiar thing happened to my words... I took F51R and replaced
> gallows characters with 'd + single quote' and crossed gallows with 'cdh'
> and then look at how close the words match those with a standard 'd'....
>
> d'sholdchy qod'chy od'chear yd'chedy
> dcheodaiin cdheody cdhody chody
> ydchody cdhey od'y cdheodar qod'y
> daiin cdhes od'ol cheody cdhy cheeey
> d'cheody qodaiin od'eey qecdhey d'aiin
> ychodaiin chod'aiin yd'chodaiin yd'ald
> ych'osar eecdhy d'cho qod'chy qotal
>
> Well that was fun wasn't it? Hmmm, what if Q's were changed to ch' too...
> then line three ends in ch'od'y. Nothing like reducing the overall character
> set to nothing, eh?
>
> John.