book: Digital fortress
author: Dan Brown
publ: st. martin's paperbacks
such a brilliant work of fiction! this is what was felt by an ordinary reader like me when i finished reading it. i guess many others like me who don't have a deep understanding of cryptography or computers must have felt the same. the brilliance i attributed to the book is not without failing to see a handful of flaws in the design of the plot and the characters in it. but the clever insertions of a liberal supply of thrilling twists and the gripping narrative makes an ordinary reader overlook such faults. except one, that is. at least i couldn't overlook that specific defect. digital fortress is the name of a supposedly malicious algorithm. but the nsa who boasts of having the ultimate code-breaking machine, transltr that has an history of cracking the most complicated and advanced codes in a matter of minutes which would otherwise have taken years fails to break it. transltr couldn't decipher the code even after hours of working on it. the explanation is that digital fortress makes use of the so called technique of rotating cleartext, which simply means that by the time the transltr guesses the correct pass-key, the decrypted data will have mutated to something else. so, in effect, the transltr succeeds in finding the pass-key, however long and complicated it be, but still can't open the code as the pass-key becomes useless by that time. and yet, the entire story revolves around the desparate search challenged by unfavourable situations for a key that is supposed to decrypt the code, as transltr has failed to solve the problem. the question is that if such a code exists, then the transltr would have actually guessed it and opened the coded data long back within a matter of minutes. because transltr's job is to actually keep guessing all sorts of passkeys-- a trial and error method. as the transltr is made up of millions of processors, it will take only a few minutes to guess even the longest passkeys which ordinary computers may take years. and that is exactly how the transltr is relevant. according to the plot, even if one guesses the correct passkey, it will be of no use as the algorithm is built up with the rotating cleartext function. then what is the use of a quest for a passkey? and how can such a key exist for such an algorithm in the first place? a corollary is that who would want to make such an algorithm that cannot be opened. who would want to make an encryption that cannot be decrypted? the flaw is so serious that the entire story is based upon it. and therefore, if the flaw is recognised, the entire series of incidents that make up the story, and thus the story itself, will cease to exist. it is hard to think that dan brown did not notice this fault. an author who designs such a thrilling page-turner with countless twists and supenseful turns failing to notice it is an implausible thing. am i the only one who think about this fault? i browsed the amazon.com reviews made by readers and also googled for a while, but i couldn't find anyone talking about this specific shortcoming in the plot, not even the most ferocious criticizers. that made me suspect if i missed some point. isn't it a flaw at all? if anybody has a good explanation, please tell me. perhaps i missed something somewhere. perhaps it's not a flaw at all and it may have some logical explanation.
to my surprise, when i browsed people's reviews online, i found many talking about more defects in the book than i percieved. but those who have found them are people with a good knowledge of cryptography and computers and even about spain where a major portion of the story takes place. perhaps my ignorance in such subjects helped me appreciate and enjoy the book more than them. but anyway, the faults they mentioned belong to the kind i mentioned in the beginning which brown's sizzling narration can compensate for to a great extent. some people accused him of being a 'google-author', one who googles for information and incorporates it into his book, leading to factual errors and discrepancies, which clearly exhibits the author's lack of first-hand knowledge of things he deals with. i too indeed found in various portions of the book some descriptions that looked like cut-and-paste information. in his da vinci code, which is the first book of his that i read, it didn't bother me much. but in this book, which is the second book of brown that i read, when i came across the same style, i couldn't just ignore the repetition.
the greatest marvel in the book to me is the meaning of the raising of his hand by ensei tankado towards the people around him when he was dying of an apparent heart attack. the meaning is revealed only at the end. the twist is simply awesome!!
Thursday, May 24, 2007
book: Digital fortress