🚀 go-pugleaf

In article <1992Aug4.180852.11447@kpc.com> hollasch@kpc.com
(Steve Hollasch) writes:
>
>    This is a bit out of the mainstream of the topics in this group, but I
>think it's still in keeping with the main idea.
>
>    I was playing around the other day with developing a transcripted
>shorthand that is fairly secure.  Obviously this will be much weaker than,
>say, crypt or the DES, but I wonder if there can be some "reasonable"
>degree of security in a written code.
>
>    One of the limiting requirements is that I want to be able to write it
>down quickly, without doing calculations for each symbol, so it would have
>to be something like an substitution alphabet.  In addition, I tailored my
>symbols to be easy to write quickly; ideally I'd like to be able to write
>at a speed comparable to (faster than?) writing standard English text.
>
>    So what I ended up with last night is basically a simple substitution
>cipher with some tricks tossed in to lightly screw up letter distribution
>(since, in my understanding, the "etaoinshrdlu" attack is very effective
>for transcribed English code).  I basically did things like encode double
>letters differently, combine some symbols (like "ng"), drop vowels
>occasionally, and the like.
>
>    So are there other tricks for developing such an encryption?  Are
>there any references for this sort of thing?  Also, does anybody know 
>how many different types of shorthand there are?

This sounds sort of like a modern version of the "nomenclator" type
of code widely used in the middle ages.  Kahn ("The Codebreakers")
has a chapter or two on them.

- Jonathan Thornburg
  <jonathan@einstein.ph.utexas.edu> or <jonathan@hermes.chpc.utexas.edu>
  University of Texas at Austin / Physics Dept / Center for Relativity
  and (for a few more months) U of British Columbia / {Astronomy,Physics}