[ietf-dkim] 1358 ssp-requirements-01 // DKIM Strict definition needed.

Douglas Otis dotis at mail-abuse.org
Mon Oct 9 11:13:04 PDT 2006


https://rt.psg.com/Ticket/Display.html?id=1358

Prior to the policy requirements, there were several supporting this  
concept of "strict".  There is a need for more than just an assertion  
that "all messages are signed."  Making an assertion that "all  
messages are signed" might mean only messages with invalid signatures  
should be introduced by services known to damage signatures.  This  
would be an incorrect assumption when dealing with commerce related  
transactions from a heavily phished domains.  The need for this added  
assertion is already found in Eric's latest SSP draft.

Two assertions are required when all messages are initially signed.   
Otherwise the partial information of "all messages are signed" may  
induce improper handling.  This would be especially true when sources  
known to damage signatures are used to enable exceptions.

To avoid improper handling two assertions must be allowed:

1 ) All messages are signed.
2 ) Services that might damage the signature are avoided.

-Doug





More information about the ietf-dkim mailing list