CrypTool is the world’s leading crypto e-learning software. Currently, the 20th anniversary of this outstanding tool is celebrated in Munich. Among the guests are some renowned codebreakers.

In my blog posts, I often mention the open-source software CrypTool. When it comes to performing codebreaking tasks, such as frequency analysis or computing the index of coincidence, CrypTool is extremely helpful. However, CrypTool can do much more. It supports hundreds of crypto functions, including AES, RSA, Diffie-Hellman, MASCs, ADFGVX, turning grilles, and transpositions. Many of these methods are visualized, which helps the user to understand what is going on. All in all, CrypTool is a great crypto e-learning and codebreaking software.

 

20 years, over 400 crypto functions

The CrypTool project started in 1998 at Deutsche Bank. The concept was developed by Bernhard Esslinger, who is today a professor at the University of Siegen. Over the years, this software grew to a more and more powerful tool. Numerous companies (such as my employer cryptovision) and non-commercial organisations provided code for CrypTool. Students developed CrypTool functionality as part of their theses.

Meanwhile, there are four CrypTool variants: CrypTool 1, CrypTool 2 (the one I usually work with), JavaCrypTool and CrypTool-Online. More than 400 crypto-related functions are implemented in these four programs.

Starting this morning, the CrypTool team is currently celebrating the 20th anniversary (yes, meanwhile it is one year too late) of their project in Munich, Germany.

Source: Schmeh

I have the honor of taking part in this event. My main contribution to the CrypTool project is that I have mentioned this software in many of my books, articles, research works and blog posts. I hope that my work has helped a little to make CrypTool more popular.

The following picture shows the five fathers of CrypTool (Prof. Arno Wacker, Jörg Schneider, Dominik Schadow, Henrik Koy, Prof. Bernhard Esslinger):

Source: Schmeh

Today, the main developer of CrypTool 2 is Nils Kopal. The next picture shows him demonstrating the software:

Source: Schmeh

 

MTC3

Another part of the CrypTool project, managed by almost the same people, is the MysteryTwister C3 (MTC3) website. MTC3 is the most popular crypto challenge portal worldwide. It provides hundreds of ciphertexts for the user to break, in four different difficulty levels (some of these challenges were provided by me). For solving a challenge, a user gets a certain amount of points. The most successful solvers are displayed in a hall of fame.

At the CrypTool anniversary celebration, four of the five most successful MTC3 participants are present. Here they are:

Source: Schmeh

Hajo Eilau (#2), Kurt Gebauer (#5), George Lasry (#1), and Armin Krauß (#4). All four are world-class codebreakers. Readers of this blog might know George and Armin, who have been mentioned many times on this blog. Hajo and Kurt are newcomers in the codebreaking scene. I have never met them before. I hope we will hear a lot from them in the future.


Further reading: The Top 50 unsolved encrypted messages: 8. The Catokwacopa cryptograms

Linkedin: https://www.linkedin.com/groups/13501820
Facebook: https://www.facebook.com/groups/763282653806483/

Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Kommentare (6)

  1. #1 Thomas
    1. November 2019

    Congratulations to the developers!

    The slide says “CrypTool is THE e-learning tool for cryptology”. As it seems, this refers to students. But what about machine learning? There have been approaches to train neural networks. For instance LSTMs (long short-term memory), a neural network technique used for working on sequential data such as in speech recognition and translation software, were feeded with ciphertext/plaintext pairs generated with Vigenere and particular Enigma settings: https://arxiv.org/abs/1708.07576v2. If I get this right, the difference to optimization approaches such as hill climing and simulated annealing is that the decryption algorithm of the particular cipher isn’t implemented inside. The neural network only learns from a great amount of ciphertext/plaintext samples. Since CrypTool provides a vast variety of ciphers, couldn’t it be used for training such a LSTM network? Is deep learning reasonably applicable in cryptanalysis or is that a wrong track?

  2. #2 bE
    2. November 2019

    Hi Thomas.

    The answer to both of your questions is Yes: CT can be used for generating training data, and deep learning is reasonably applicable in cryptanalysis.

    At the CrypTool conference yesterday, we had a talk from G. Lasry using e.g. hill climbing against pre-computer ciphers like ADFGVX or SIGABA. We also had a talk from A. Gohr, who successfully used deep learning to attack the modern, round-reduced NSA cipher Speck32/64 (see https://eprint.iacr.org/2019/037).

    We are planning to enhance CrypTool in this direction. Please send an email to bernhard.esslinger@uni-siegen.de, if you want to participate in that goal in one way or another.

    Best regards, BE

  3. #3 Marc
    2. November 2019

    @bE
    Do you also plan to implement octagrams as fitness function like Jarl Van Eycke did in order to solve the latest bigram challenge ont this blog?

  4. #4 Nils Kopal
    Krefeld
    5. November 2019
  5. #5 Nils Kopal
    Krefeld
    5. November 2019

    Concerning the question about “8grams”.
    Yes, we have this in mind, but this has not now top priority since we are working on many other things right now.

    With the help of the “CrypToolStore” it should be relatively easy to offer additional (bigger) language statistics which the users then may download when they need it (for more sophisticated cryptanalysis).

  6. #6 bE
    5. November 2019

    Concerning the question about “8grams”.

    @Jarl Van Eycke

    Congratulations also from me to you and Louie for breaking Klaus’ “bigram 1000 challenge” (http://scienceblogs.de/klausis-krypto-kolumne/2019/10/27/bigram-1000-challenge-solved-new-world-record-set/).

    Would you be willing to support integrating your algorithm and your 8-gram statistics into CT2?
    If so, please don’t hesitate to contact me via email:
    bernhard.esslinger@gmail.com