On Thu 26/Dec/2024 04:00:19 +0100 John Levine wrote:
It appears that G.W. Haywood <[email protected]> said:
Hi there,
Not wanting to throw any spanners in any works here, and assuming that
it isn't already implicit, is it worth suggesting somewhere in the RFC
that aggregate reports from one entity to another for any given time
period should themselves (i.e. for that unique set of [entity, entity,
time period]) be unique?
Amazingly, I agree with Doug here. Amazon has a zillion mail servers and
different groups send different reports. The report_id fields are different
so you know the reports are different. If your database has trouble
adding them together, it's not much of a database.
My understanding is that Ged was inclined to suspect some kind of bug at the
report generator's.
Reporters are doing you a favor by sending you reports, so it would not be
a good idea to invent new rules. A reasonable response would be it's
too hard to send reports that follow the new rules, so we won't send any
reports at all.
A new rule doesn't have to be hard. A report-generator-id, a UUID created once
on installation would suffice.
Rather than following new rules, I guess most reporters will just ignore them.
I'm waiting for our schema to be published at IANA[*] before referencing it in
the reports I generate. After the schema is published, adding a new field will
be neither easy nor reasonable.
Best
Ale
--
[*] https://www.iana.org/assignments/xml-registry/xml-registry.xhtml
_______________________________________________
dmarc mailing list -- [email protected]
To unsubscribe send an email to [email protected]
