Interesting. I wasn’t impressed by his Fermat’s last theorem book which was basically purely a biography of various mathematicians and contained very little actual information. Is this one better?
As it happens I found Applied Cryptography extremely readable (albeit out of date), so in this case I’m happy to recommend a textbook.
Assuming that rank-biserial is referring to the same book I have on my shelves with the title of “The Code Book”, here’s a brief breakdown of what’s in it, ignoring merely-historical stuff like biographical titbits about Alan Turing:
“The cipher of Mary Queen of Scots”: introduces simple “classical” ciphers—transpositions, simple-substitutions, and the like—and says a bit about breaking them via frequency analysis. The titular cipher isn’t quite a simple substitution cipher but isn’t far off. It was broken, presumably basically by frequency analysis; Singh doesn’t say much about how.
“Le chiffre indechiffrable”: the Vigenere cipher, other polyalphabetic substitutions, homophonic substitutions, the “Great Cipher” of Louis XIV (repesenting syllables by numbers), Babbage’s approach to attacking Vigenere-enciphered text (look for repeating sequences in the ciphertext; their spacing is probably a multiple of the key length; the idea is usually credited to Kasiski, because he published it and Babbage didn’t), some 19th-century fiction about ciphers, the Beale ciphers.
“The mechanisation of secrecy”: wartime ciphers: ADFGVX, the Zimmerman telegram (no information about the actual cipher used), random one-time pads (motivated by showing how in practice it might be possible to decipher a Vigenere-ciphered message where key and message are the same length, if the key and message are both ordinary text rather than random garbage), cipher machines with particular reference to Enigma.
“Cracking the Enigma”: Rejewski’s cycle-length-based approach (described in reasonable detail), his bombes, transfer to Bletchley when extra wheels and plugboard cables made decryption too onerous for the Poles, Turing’s crib-based approach (described in reasonable detail) and his bombes, vague stuff about other WW2 ciphers.
“The language barrier”: Navajo code-talkers, Egyptian hieroglyphics and the Rosetta stone (Young, Champollion), Linear B (Ventris, Chadwick). Reasonably concrete about these.
“Alice and Bob go public”: DES and Feistel ciphers (bizarrely, the latter described in text but no diagram), Diffie-Hellman key exchange (actual algorithm given), public-key encryption, RSA (details left to an appendix), a bit about the earlier discovery of the same ideas at GCHQ.
“Pretty good privacy”: PGP, symmetric session key shared using asymmetric cryptography, legal nonsense around crypto exportation etc., key escrow, certification authorities.
“A quantum leap into the future”: handwavy discussion of breaking ciphers with quantum computers (mostly at the “quantum computers do exponentially many things in parallel” level), quantum cryptography (with actually a reasonable level of detail).
So, like his FLT book, it’s more historical than technical. If you read it and pay attention throughout you will learn some technical details, but e.g. you won’t learn anything about how modern symmetric ciphers are designed and broken, or what distinguishes good cryptographic protocols from bad ones, or implementation-level details like side-channel attacks, or any public-key methods beyond RSA.
If you went into Fermat’s last theorem looking for an intro to elliptic curves you were out of luck but as a book about why maths is awesome it is great—precisely, I think, because of that emphasis on mathematicians biographies. Reading it was one of the pivotal moments in my life, setting me on the path that lead me to a maths PhD (and incidentaly lesswrong too).
Code by Simon Singh is a nice, low-effort intro to cryptography.
Interesting. I wasn’t impressed by his Fermat’s last theorem book which was basically purely a biography of various mathematicians and contained very little actual information. Is this one better?
As it happens I found Applied Cryptography extremely readable (albeit out of date), so in this case I’m happy to recommend a textbook.
Assuming that rank-biserial is referring to the same book I have on my shelves with the title of “The Code Book”, here’s a brief breakdown of what’s in it, ignoring merely-historical stuff like biographical titbits about Alan Turing:
“The cipher of Mary Queen of Scots”: introduces simple “classical” ciphers—transpositions, simple-substitutions, and the like—and says a bit about breaking them via frequency analysis. The titular cipher isn’t quite a simple substitution cipher but isn’t far off. It was broken, presumably basically by frequency analysis; Singh doesn’t say much about how.
“Le chiffre indechiffrable”: the Vigenere cipher, other polyalphabetic substitutions, homophonic substitutions, the “Great Cipher” of Louis XIV (repesenting syllables by numbers), Babbage’s approach to attacking Vigenere-enciphered text (look for repeating sequences in the ciphertext; their spacing is probably a multiple of the key length; the idea is usually credited to Kasiski, because he published it and Babbage didn’t), some 19th-century fiction about ciphers, the Beale ciphers.
“The mechanisation of secrecy”: wartime ciphers: ADFGVX, the Zimmerman telegram (no information about the actual cipher used), random one-time pads (motivated by showing how in practice it might be possible to decipher a Vigenere-ciphered message where key and message are the same length, if the key and message are both ordinary text rather than random garbage), cipher machines with particular reference to Enigma.
“Cracking the Enigma”: Rejewski’s cycle-length-based approach (described in reasonable detail), his bombes, transfer to Bletchley when extra wheels and plugboard cables made decryption too onerous for the Poles, Turing’s crib-based approach (described in reasonable detail) and his bombes, vague stuff about other WW2 ciphers.
“The language barrier”: Navajo code-talkers, Egyptian hieroglyphics and the Rosetta stone (Young, Champollion), Linear B (Ventris, Chadwick). Reasonably concrete about these.
“Alice and Bob go public”: DES and Feistel ciphers (bizarrely, the latter described in text but no diagram), Diffie-Hellman key exchange (actual algorithm given), public-key encryption, RSA (details left to an appendix), a bit about the earlier discovery of the same ideas at GCHQ.
“Pretty good privacy”: PGP, symmetric session key shared using asymmetric cryptography, legal nonsense around crypto exportation etc., key escrow, certification authorities.
“A quantum leap into the future”: handwavy discussion of breaking ciphers with quantum computers (mostly at the “quantum computers do exponentially many things in parallel” level), quantum cryptography (with actually a reasonable level of detail).
So, like his FLT book, it’s more historical than technical. If you read it and pay attention throughout you will learn some technical details, but e.g. you won’t learn anything about how modern symmetric ciphers are designed and broken, or what distinguishes good cryptographic protocols from bad ones, or implementation-level details like side-channel attacks, or any public-key methods beyond RSA.
If you went into Fermat’s last theorem looking for an intro to elliptic curves you were out of luck but as a book about why maths is awesome it is great—precisely, I think, because of that emphasis on mathematicians biographies. Reading it was one of the pivotal moments in my life, setting me on the path that lead me to a maths PhD (and incidentaly lesswrong too).