If he's using Denis Bider's implementation of aesphm, it uses AES in CFB mode, which does require an IV. I believe Denis' implementation generates the IV from a 256 Hash of a randomly generated IV seed.
However, I don't believe that implementation takes the IV in as a parameter. Instead, it buries the IV_SEED in the encrypted text and then pulls it back out during the decryption. If you look at the .Net implementation of Rijndael, it also requires an IV to be passed to the encryptor and decryptor objects. David -----Original Message----- From: Shawn Masters [mailto:[EMAIL PROTECTED] Sent: Thursday, February 05, 2004 6:03 AM To: 'Russell Robinson'; [EMAIL PROTECTED] Subject: RE: AES decrypting IV AES does not have an IV associated with it. IV's (Intial Vectors) are part of certain modes of using a block cipher. What mode are you using? Also remember that Crypto++ is a toolkit. It gives you the base primitives to be put together, and you can then make just about anything you need. In common cases though it may have a set of primitives tied together already for doing common tasks. 73, Shawn > -----Original Message----- > From: Russell Robinson [mailto:[EMAIL PROTECTED] > Sent: Thursday, February 05, 2004 2:50 AM > To: [EMAIL PROTECTED] > Subject: AES decrypting IV > > Hi, > > Can someone confirm that when using AES, the IV that was used for the > encryption must be sent to the decryptor in addition to the > ciphertext> > > Implementations of RC4 I've seen, for example, automatically append > the IV to the ciphertext. > > It appears AES in Crypto++ doesn't do this so I have to append it > myself. > > Correct? > > Thanks. > > -- > Russell Robinson (mailto:[EMAIL PROTECTED]) > Author of Tectite (CRM and Licensing for Software Developers) Download > your free CRM from: http://www.tectite.com/
