You are correct that a receiver can freely decide not to use MARID.
Thanks for that.
But if a sender has no expectation and a receiver has no expectation to
understand the sender's assertion, then how well will MARID work? Is
If someone decides not to use it then it will not work for them.
A Publisher cannot require that a receiver pay any attention to their
record. Having one identity or 8 *does not* change this. If a producer (or
consumer) finds that limiting their assertion (or evaluation) to one
identity works for them - that's what they'll do. Others may choose
otherwise.
Is this much different than just publishing the input docs as they are
and
letting senders and receivers pick their favorite?
Yes it is, I believe that we're tasked with developing a "DNS-based
mechanism for storing and distributing information associated with [...]
authorization." Not sitting in judgement on current (and all possible
future) MARID dependent "anti-spam" proposals, and refereeing some ghastly
religious war.
In order to accept as much legitimate email as possible using MARID
while thwarting spam, receivers have to understand the assertions of
the sender.
Perhaps, but not necessarily *all* of the senders assertions. Why would
they? We should remember that the charter directs us to come up with a
solution (to the MARID problem) which "should be generally useful". That's
hard. But that's what we're tasked to do.
A common strategy for dealing with this kind of thing is to pick a sensible
default, but leave your options open. This is not rocket science, however
much the rocket scientists out there try to make it look as if it is.