CrypToolProject
can't reproduce Fialka test vectors
According to the comments the Fialka test vectors from FialkaTest.cs are generated with the Fialka Simulator version 5.16 by Vyacheslav Chernov, however I cannot reproduce them.
For the first test vector "TESTTESTTESTTEST" with the simulator I get "PHAFR2EMWW8B8XR7" and with CrypTool-2 I get "PH7FR2EMWW8B8XRA" so the 7 and the A are swapped.
I tried all possible combinations of settings for the M-125 and only for the 3K/6K rotor series and DE keyboard I get matching results. For the 3K/6K rotor series and CH/PL keyboards the 7 and A are swapped, and for the 5K rotor series I get totally different results.
Did you use a different version of the simulator? The filename of my version says "M125v5_16eng.txt" but the file properties say version 5.0.4.0 and date modified 24/10/2014 11:43. Or did you maybe use a non-identity punchcard?
A few months later I took another look at this and I found that in the FialkaTest.cs below the advanced test cases with layout = GDR = DE there is a comment "!!! Fialka Simulator v5.16 by Vyacheslav Chernov BUG - KEYBOARD: A/8; PAPER OUT: 7/8; INPUT: 8/A; OUTPUT: A/8" which seems to be related the swapping of 7 and A that I found. Strangely enough the cases with laydout = GDR = DE are the only cases for which I get the same results from CrypTool and the Simulator.
In the same file there is also a comment about the 5K rotor series that probably explains my problems with those: "!!! Actually there is a bug in the Fialka Simulator v5.16 by Vyacheslav Chernov, for 5K rotor series, the wiring for core side 1 and 2 are exchanged. !!! We used the wiring (and calculated the wiring for flipped cores) presented in http://www.cryptomuseum.com/crypto/fialka/m125_3/hu.htm.".