On Wed 01/Sep/2021 14:34:52 +0200 Brotman, Alex wrote:
It feels like folks would prefer that the subject be required to be of a
specific format to better enable duplicate report processing. Do I understand
that correctly?
So that would be:
If a report generator needs to re-send a report, the system
MUST use the same filename as the original report.
Right. That makes it easier to recognize duplicates.
And:
The RFC5322.Subject field for individual report submissions
MUST conform to the following ABNF:
The filename is defined like so:
filename = receiver "!" policy-domain "!" begin-timestamp
"!" end-timestamp [ "!" unique-id ] "." extension
So one may ask why the unique-id is optional in the filename while msg-id is
mandatory in the Subject:?
dmarc-subject = %x52.65.70.6f.72.74 1*FWS ; "Report"
%x44.6f.6d.61.69.6e.3a 1*FWS ; "Domain:"
domain-name 1*FWS ; from RFC 6376
%x53.75.62.6d.69.74.74.65.72.3a ; "Submitter:"
1*FWS domain-name 1*FWS
%x52.65.70.6f.72.74.2d.49.44.3a ; "Report-ID:"
msg-id ; from RFC 5322
To answer that question, we need to establish what is the purpose of
standardizing the Subject: line in aggregate reports, as it seems to duplicate
the information already present elsewhere. There are two difference between
Subject: and filename:
1. Subject: is a header field while filename is only a parameter. As such, it
is subject to RFC 2231 encoding, which some applications have difficulties with.
2. Subject is prominently displayed by email clients.
What do we target, ease of scripting or readability? For the former, we'd be
better off defining a new header field and specifying how to encode it in case
of IDNs. Why is it domain-name here and policy-domain there? And why does
domain-name mention RFC 6376 and not RFC 8616?
If we target readability, instead, msg-id is just clutter. It's unreadable,
its relation with the unique-id is not specified, and the Message-ID: is a 1st
class header field, so what is the reason to repeat its value, often longer
than what the displayed Subject: can fit?
BTW, the relation between the unique-id of the filename and the report_id
inside the XML is not specified too. In the reports I send and in some of the
ones I receive those three IDs coincide, but perhaps the spec should say
something about that, shouldn't it?
Best
Ale
--
_______________________________________________
dmarc mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dmarc
