Jarlve wrote:smokie treats wrote:And what I came up with at 0.37 MIPS for 10 hours. The Acer is locked up right now.
Here is a current build of AZdecrypt 1.15 with the much faster and improved row bound solver: https://drive.google.com/open?id=1YUKgz ... hHCc0yE25y
It needs the base 1.14 install to work. Also put all the files of N-gram folder in this zip in your existing N-gram folder and if Windows ask to overwrite or replace do so.
Let me know how it goes. I still need to look into your analysis, it looks interesting.
Thank you very much.
Here are the final results for American Short Stories, matrix posted above.
P16, P32, P48 pivots in cleartext not transposed
127 messages with one pivot
17 messages with two pivots
P16, P32, P48 pivots in cleartext drafted LRBT diagonally P16
54 messages with one pivot
1 messages with three pivots
P18, P36, P54 pivots in cleartext not transposed
129 messages with one pivot
11 messages with two pivots
6 messages with three pivots
P18, P36, P54 pivots in cleartext drafted LRBT diagonally P16
133 messages with one pivot
11 messages with two pivots
1 messages with three pivots
I do not think that I have an error with the P16, P32, P48 pivots in cleartext drafted LRBT diagonally P16 test. Maybe the lower pivot counts because the letters must be tied to each other in an ngram for each pivot pair, like OO, A_A, and E__E, and line up into a pivot, whereas for the other tests the letters do not have to be tied to each other.
