On Mon 06/Dec/2021 20:57:22 +0100 Brotman, Alex wrote:
Hello folks,
We received a ticket about the unique IDs that are included with reporting:
--------------------------
unique-id, msg-id, and report_id are loosely defined.
The spec needs to say that they are semantically the same thing and must have
the same content.
In addition:
Message-ID should also be based on unique-id if possible (for example,
msg-id@domain-name),
make optional the final part of the Subject:, [Report-ID: msg-id] (like
the filename).
--------------------------
I agree, the references should be made to the point to the same identifying
string definition. Today, the msg-id is required via the Subject, report_id is
required in the XSD, and the unique-id is optional in the filename of the
attached report. Proposed is adding a suggestion to use the same ID attached
to the Message-Id header. Thoughts about making all required or optional?
I think some should be mandatory while others should be optional. Two examples:
*filename*:
Some record producers issue multiple messages for the same period in order to
overcome size limit restrictions. In that case the unique-id is the only means
to recognize duplicate filenames without digging into their content. I'd say
this has to be required.
*Subject*:
This field is only relevant for humans who see the reports in a message list
pane. In that case, it is unavoidably truncated and duplicated messages are
more easily spotted by other means. Rather than overturning RFC 7489's rule,
I'd make it optional.
Best
Ale
--
_______________________________________________
dmarc mailing list
dmarc@ietf.org
https://www.ietf.org/mailman/listinfo/dmarc
