[Noisebridge-discuss] [CONLANG] Why do so many crackpots go for "crypto"?

[Sai Emrys]

FYI, I crossposted this to:

a) CONLANG, a list for people who make languages (has people w/ high
linguistics knowledge)

http://listserv.brown.edu/archives/cgi-bin/wa?A1=ind1006c&L=conlang#6

b) Noisebridge, a list for hackers associated with the SF hackerspace
named Noisebridge (has people w/ high coding and crypto knowledge)

https://www.noisebridge.net/pipermail/noisebridge-discuss/2010-June/thread.html#14624

There's been good discussion on both, so I'd suggest people interested
read the other list's posts.


A couple other things to consider for the decrypt:
* maybe F and/or Q are meta characters (e.g. capitalization, quotes,
space, punctuation)?
* maybe it's using some sort of shorthand / contractions?
* for the main body, are linebreaks significant?
* are there spaces in the cyphertext? I vaguely think there are but
I'm suspecting myself of paredolia.
* are there digraphs, or mono<->digraph mappings (e.g. maybe 'th' is a
single character)?
* is it really bad English, or English with some embedded Arabic words?

I'm hesitant to actually give this guy enough credit to say that any
of the above are actually *plausible* - they're mostly things that
would require a little more cleverness than what is more likely to be
a more straightup crackpot confusion of character substitution for
crypto. But maybe.

FWIW, FQ being both extremely common and asymmetrical bigrams seems to
be strong argument in favor of it being =TH. But I don't have any good
idea for how to handle the weird stretches where there're a whole
bunch of 'em close together. They seem too high percentage.

- Sai