Thanks for your reply... But I am looking for more specific answer. >From the README I get: Key=Key[Block# % 64] IV=MD5(Block#, Key[64]) EncryptedBlock=AES_CBC(Key, IV, Block) And I need to know, why it is unbreakable... For example: 1. Where does the magic 64 come from? 2. Is MD5 remains safe after recent developments in this area? 3. Is the usage of the same key over and over is "good enough" still? The GPG part is unrelated... Let's say we load the keys from external media. Thanks, Alon. On Mon, Mar 30, 2009 at 5:59 PM, Boyd Waters <bwaters@xxxxxxxxxxxx> wrote: >> I need to convince someone to use it, and it would be handy to >> provide some theoretical paper done based on loop-aes. > > I believe that the strength of loop-aes is in the robustness of the > implementation, and the great write-up (README) that Jaari provides, > showing that strong key management is possible (use GNU PG) --- > > In short, the overall cryptoSYSTEM is the best available. > > Most other AES implementations are less attractive (IMHO) because of > the weakness or the complexity of their key management -- how it > interacts with real users. > > I don't know how to turn this into a mathematical argument: all > correct AES implementations may look the same, mathematically. But > user interaction and key management can be difficult. loop-aes keeps > it simple. > > (Oh, I suppose that multiple keys, different key per block is a strong > point that has a mathematical proof.) > > - boyd > > Boyd Waters > Scientific Programmer > National Radio Astronomy Observatory > New Mexico, USA > Earth > - Linux-crypto: cryptography in and on the Linux system Archive: http://mail.nl.linux.org/linux-crypto/
