When switching to SHA-2 message digest for S/Mime there is a problem that it is not always known if the other side will understand it. After all SHA-2 is defined for S/mime in RFC 5754, but not all clients support it.
So I am looking for a way to produce universally accepted signed messages.
According to RFC3851 you can specify a list of message digest algorithms. It looks like you would have two Signatures SignerInfo blocks on the message then (from the same key).
Is this in practice actually done and has anybody compatibility experiences with this? With BC I guess you would add multiple signerInfoGenerators?
Chief Architect (R&D), SEEBURGER AG, Germany
Yes, you'd need to add multiple signerInfoGenerators.
I think you'd need to make sure you used different public keys for the signatures though. Apart from possible issues with generating lower security signatures with a weaker hash with the same key, I don't think many clients would do anything sensible if two signers had the same SignerID (or put another, while there might be a SHA-1 signer information object present, there's always the chance the "SHA-1 only" client will fetch the SHA-2 based signer information by mistake, not realising there is another one present).
On 19/10/16 03:02, Eckenfels. Bernd wrote:
|Free forum by Nabble||Edit this page|