Provided that the body and signed headerJim Fenton – Aug 07, 2008 12:45 PM PST
Provided that the body and signed header fields are not altered in the forwarding process, DKIM signatures should be expected to survive. This specific use case was considered in the design of DKIM and is an advantage of signature-based authentication.
What are the other technologies of choice?Jim Fenton – Aug 08, 2008 12:17 AM PST
I don't really understand the alternatives you mention: Handling exceptions at the MTA is very much like forwarding, and some sort of MX record hack seems even less likely to modify the message in a way that would break the signature. Am I missing something?
Provided that the body and signed header fields are not altered in the forwarding process, DKIM signatures should be expected to survive. This specific use case was considered in the design of DKIM and is an advantage of signature-based authentication.
Jim, thank you for clarifying that point. I appreciate your correction.
It also seems to me, though, that forwarding may not be the technology of choice for these ISPs.
I don't really understand the alternatives you mention: Handling exceptions at the MTA is very much like forwarding, and some sort of MX record hack seems even less likely to modify the message in a way that would break the signature. Am I missing something?
Yes, the legislator may demand a movement, not of the pointer but of the object, the mailbox.
Get your own domain name and you have mail portability.