The CONDI (CONsecutive DIgraphs) cipher was introduced by G4EGG (Wilfred Higginson) in the Sept-Oct 2011 issue of the ACA's Cryptogram magazine. The cipher preserves word divisions, and is simple to describe and encode, but it's surprisingly difficult to crack even with a computer.
The encoding steps are:
(1) Start with a key word or phrase, say CRYPTOGRAM.
(2) Form a key alphabet by removing from the key duplicated letters except for the first occurrence, and appending in alphabetical order all letters than do not occur in the key. Our example would produce this key alphabet:
CRYPTOGAMBDEFHIJKLNQSUVWXZ
(3) Number the key alphabet starting with 1.
1 2 3 4 5 6 7 8 9 10 11 12 13
C R Y P T O G A M B D E F
14 15 16 17 18 19 20 21 22 23 24 25 26
H I J K L N Q S U V W X Z
(4) Choose an arbitrary starting number between 1 and 26, say 10, encode the first letter of your plaintext by
moving 10 places to the right from the letter's position in the key alphabet. If the first letter were say 'o' then the letter 10 places to the right in the key alphabet is J, so 'o' would be encoded as 'J'. If you move past the end of the key alphabet you wrap back to the beginning. For example if the first letter were 's' then counting 10 places would bring you around to T.
(5) Use the position of the previous plaintext letter as the number of places to move to encode the next plaintext number. If you have just encoded an 'o' (position 6) , and you now want to encode say 'n', then you move 6 places to the right from 'n' which brings you to X.
(6) Keep repeating step 5 until all letters are encoded.
Decoding is the reverse of encoding -- you move to the left instead of to the right.
The length suggested by the ACA for the CONDI cipher is 100-200 letters.
Here's an example using the key alphabet and starting number above:
plaintext: On the first day I got lost.
ciphertext: JX WNZ XRKVZ JND L UFD VWCZ.