- From: Amir Herzberg <AMIR@newgenpay.com>
- Date: Wed, 27 Jun 2001 10:59:30 +0300
- To: "'Frederick J. Hirsch'" <hirsch@zolera.com>, xml-encryption@w3.org
Fredrick said, > The XML Encryption document specifies that DigestMethod and > DigestData may be > specified > within the CipherData element to provide integrity. > http://www.w3.org/Encryption/2001/Drafts/xmlenc-core/#sec-CipherData > > I assume the same result could be achieved by not including > these elements, but > using an XML Signature > on the content to be encrypted, and then encrypting both the > content and the > signature. This would have the additional cost of maintaining > the keys for > signing and implementing XML signatures. Correct. And, the spec does not prohibit this (DigestData is optional). > It would have the > benefit of providing > stronger integrity than a simple hash. The hash by itself does not provide protection. However, by including it, a signature can sign the hash (in addition or instead of signing the ciphertext), providing strong authentication and integrity protection. > > I propose we leave this up to the application rather than > defining the digest > elements as part of CipherData As these elements are optional, we do not exclude the other method. If we remove them, we cannot use this more efficient and flexible method. > > Alternatively we can leave the optional > DigestMethod,DigestData elements in the > schema but suggest that stronger (source) integrity be obtained with a > signature. The stronger integrity and authentication requires of course signatures, but the signatures can be applied to the hash of the plaintext (DigestValue). > > < Frederick > > hirsch@zolera.com >
Received on Wednesday, 27 June 2001 03:55:47 UTC