First - the overall point of this is to be able to better report on
authentication headers. If you are only doing dkim then the dkim reporting
is probably a better bet than this one. This is meant to be the next step
for marf for reporting authentication failures in a uniformed way.
For 3.2.1 delivery result - you can default to other but the preference
would be to keep it as a MUST
I think you're misunderstanding what a standard is. It says, if you want
to interoperate with other people, here's how you do it. No IETF
standard can ever force people to do anything they don't want to.
So if it says you MUST return a delivery result, what you're actually
saying is that if you're not going to send a delivery result, don't bother
to send anything. You would probably prefer to get some reports than no
reports, so MUST is utterly counterproductive.
I also note it is fairly common for people simply to ignore parts of
standards they think are silly, so no matter what you say, people will
send reports with only DKIM info for DKIM failures, only SPF info for SPF
failures, and no delivery results. Experience suggests that successful
standards are written to maximize the likelihood that people will
implement them as written, so you don't write requirements that are likely
to be ignored.
It appears we all agree that some things in the draft are just wrong, like
the Message-ID, so this draft needs to recycle at least once more.
Regards,
John Levine, [email protected], Taughannock Networks, Trumansburg NY
"I dropped the toothpaste", said Tom, crestfallenly.
_______________________________________________
marf mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/marf
