The Code Book

A book written by Simon Singh. Published by Doubleday in 1999. It is about the history of secret codes, cryptography, and other methods of attempting to communicate in secrecy.

I just started reading this book. I'll add more once I finish. I watched Mr. Singh give a talk about the book on C-SPAN. It sounded interesting after having read Cryptonomicon.

Update: Wow! It was a damn good book. I highly recommend it for anybody that is interested in the history of cryptology. It is far from a complete history and it does tend to lean heavy on what the British have done. It isn't a textbook that explains all the math involved with each system, but it was much easier to read than a textbook. It tells the story of the men and women involved in some of the important steps in cryptology's development. Similar to a PBS or History Channel documentory. (Mr. Singh's day job is writting those types of shows.) It also has an excellent set of reference's for further reading in the back.

It also includes a contest, The Code Book Cipher Challenge. The first person to decrypt all 10 steps will win $15,000. More info can be found at http://www.4thestate.co.uk/cipherchallenge.

The Code Book

Singh, Simon, (C)1999 Doubleday, New York, NY

The Code Book is great coverage of the history of cryptography, from the initial codes written by Mary, Queen of Scots to modern public/private algorithms and even theory on Quantum Cryptography, a modern-day holy grail of sorts. It explores the flaws in early methods of cryptography, as well as the employment of foreign and largely unstudied languages (Navaho being the prime example) for communication that has proven very indecipherable. The ride through the museum of covert communications continues as the book covers the WW2 era German Enigma machine, the story of which is full of treachery, betrayal and adventure. It covers some of the advancements made by Horst Feistel towards the creation of the DES cipher, which, while limited by the private-key problem, became widely popular as a far more secured cipher than the simple substitution ones. Public-Private key encryption, which is new and considered far more secure, has continued to evolve as computers become more sophisticated. Quantum cryptography, in which computers find every possible state for a problem, and magically approve of the right one, sounds like the fable of the sword from the lake. It is purely theory, never having actually been proven in practice.

The initial ciphers described in the book are those used by Mary, Queen of Scots. She used an inferior cryptography method, a simple substitution cipher. Several documents proving her plot to overthrow a neighboring government were written in this cipher. It was inevitably cracked and she was eventually beheaded due to her proven guilt. Simple substitution ciphers can be easily broken by looking for the most common symbol, and mapping it to the most commonly used letters, like the letter "e".

Fast-forward to the Second World War. The Germans have constructed a nearly-unbreakable rotating cipher system. The book details how amongst many bright colleagues, Alan Turing, a very intelligent British mathematician working for Britain's equivalent of the NSA (the GCHQ) discovered how to develop a electro-mechanical machine that would very quickly try and deduce logical combinations of ciphers. The Turing machine became a famous precursor to the modern computer; in fact, everything in the information age could be accredited (at least in part) to his discovery.

Public key encryption solves the problem of key distribution. It basically allows a person two different keys: a public, freely-distributed key, and a private secured key. The private key only deciphers, while the public key only enciphers. A metaphor used in the book is one of padlocks, keys and the US Postal Service. Bob mails a box with a unsecured padlock in it to Alice. Alice writes Bob her letter, secures the box with the padlock and mails it to Bob. Bob then opens the box with his key. Public-key encryption is right now the best system available for securing documents transmitted over public networks (like the Internet).

Currently, however, encryption distribution around the world is in trouble. The United States Government has deemed cryptographic software and algorithms an armament, no different than a AK-47, rocket launcher or ICBM. Therefore, it is currently illegal to export software (for free or profit) that supports keys larger than 40 bits, which at this point is very easy to crack. This is a problem as it has slowed development of cryptographic software as collaboration with developers living outside the US (and there are many, living largely in Europe, Asia and elsewhere) is made extremely inconvenient. Most strong, free cryptographic software is developed in Finland at the moment, as there are no restrictions on exporting (or importing) strong cryptography from Finland or importing it to the US. Phil Zimmerman, author of the first usable cryptographic software, is a citizen of the United States. He ran into a bunch of legal trouble with the United States Government when he distributed his software on an Internet newsgroup, theoretically allowing downloads from all over the world. There were many legal battles, and the issue of whether or not he was truly exporting or not was raised many times. It is amusing that within the US, however, as long as the government considers encryption a munition that it will be protected under the Second Amendment.

Quantum Cryptography has long been regarded as a holy grail of sorts; both in the fields of cracking existing codes and forming unbreakable new ones. Because this book details mostly the encryption-related items, this is what I will detail myself. Quantum cryptography involves the theory that matter can exist in two simultaneous states and hat by manipulating the way the dice roll, if it is at all possible, would allow for cryptography that would be completely unbreakable, with no significant flaws.

One way to encrypt a document in a "quantum" fashion is to trap a wave of light (again, not said to be possible), and get it stuck in a certain polarization when you do. For a given number of detectors, it's not possible to completely determine the phase of each wave of light.