On October 8, 2021 4:18:45 PM UTC, Dave Crocker <[email protected]> wrote:
>On 10/8/2021 9:09 AM, Scott Kitterman wrote:
>> So originator includes From and Author and signs both.  Then the mediator 
>> (e.g. MLM) minges From and signs again.  Receiver checks DMARC and it 
>> passes.  Then receiver sends feedback to both Author and From domains?
>
>The purpose of the Author field is to retain some information that 
>presumably won't get modified.  Whether to actually 'believe' that 
>information is a different matter, just as it is for all other header 
>fields.  And let's be clear that including a field in a DKIM signature 
>does NOT validate its contents.
>
>DMARC adds to the semantics with its definition of alignment. It's part 
>of DMARC, not DKIM.
>
>So it's certainly reasonable to include the Author: field in the set 
>that produce the DKIM signature, but that inclusions does not have any 
>semantic other than it didn't get changed since the signing.  Data 
>integrity is nice but is quite different from validation.
>
>Since you are pressing the concern, perhaps you could characterize what 
>danger/threat and what meaningful protection against it you are looking for?

I don't have one yet.  I'm trying to make sure I understand the proposal.  

I completely agree about integrity versus validation.  Which fields to sign is 
only relevant here because a proposal to not sign From (which it had now been 
clarified this wasn't) would require an incompatible DKIM change.

Scott K

_______________________________________________
dmarc mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dmarc

Reply via email to