The new version of DMARC took them out since nobody used them.
That's a shame, but thanks for the tip. I thought it was a legimitately good
idea, but it does also add new dependencies on LWP that aren't stock on some
OSes.
I suspect the problem is that in a lot of orgnizations the people who run
the web servers and the people who run the mail are siloed and getting a
web server set up to accept DMARC repports is more trouble than it's
worth, particularly if the alternative is a mailbox on a server they
already run, or pointing it at one of the many analysis services.
Well, if you remember the way this thread all started, it started with
bombarding a known mailbox provider that doesn't always follow standards for a
nonexistent user.
Since some mailbox providers are big black boxes that nobody can say how they
work, it's reasonable to assume that you shouldn't be sending them mail that
will bounce, lest you run afoul of some of their other undocumented
rate-limiting, which gives you no feedback even if you sign up for their
postmaster service.
Sure, but if some dimbulb puts a bad address in their DMARC record, I'm
not inclined to spend a lot of time worrying about it.
Also, as an responsible admin, I *should* be looking at bounces, but why
should I have to keep seeing bounces when I'm trying to send someone
something that they implicitly asked for but are too clueless to make
work?
See above. Our time is limited, flushing bounced DMARC reports to
/dev/null is cheap and easy. As I have often noted, the total bandwidth
all of our mail users is tiny compared to a typical cat video. It's just
not worth worrying about.
R's,
John
_______________________________________________
mailop mailing list
[email protected]
https://list.mailop.org/listinfo/mailop
