Re: Password derivation

Hi Christian,

I think it can be incorporated in the same manner as the other algorithms.
That is, you just specify the identifier of a password-based encryption
algorithm to the EncryptionMethod element and include a KeyName element or
something in the KeyInfo element in order to let a decryptor know a
password for decryption.  I'm not sure whether we should include such
algorithms in the spec, though.

Thanks,
Takeshi IMAMURA
Tokyo Research Laboratory
IBM Research
imamu@jp.ibm.com



From: Christian Geuer-Pollmann <geuer-pollmann@nue.et-inf.uni-siegen.de>
      @w3.org on 2002/01/22 01:25

Sent by:  xml-encryption-request@w3.org


To:   Joseph Reagle <reagle@w3.org>
cc:   xml-encryption@w3.org
Subject:  Password derivation



Hi Joseph,

in [1] section 4.2.10 there is stated:

"4.2.10: Password derivation: Unclear what it means, strike from
requirements."

I guess I have an idea: Currently, I'm implementing an
XML-Signature/Encryption-based KeyStore (for all the non-Java-Folks, a file
which collects encrypted Keys, Certificates etc). The keys which are used
in EncryptedData elements must be derived from user-supplied passwords.
This means that a mechanism like the one defined in PKCS12 is used to
derive a literal key from a used-supplied passphrase (or in Java lingo, to
create a byte[] key from a char[] passphrase).

Question: How could this be incorporated into XML Encryption?

Example: I want to use #kw-aes256 for wrapping an RSA private key, and the
AES 256 bit key should be derived from a user-supplied pass phrase. Do I
simply say nothing and let the application decide where to get the AES key
from?

Christian

[1] W3C XML-Encryption Minutes, Boston, MA, 01 March 2000
    http://www.w3.org/Encryption/2001/Minutes/0103-Boston/minutes.html