On Friday May 14, roy(_at_)gnomon(_dot_)org(_dot_)uk wrote:
"Neil" == Neil Brown <neil(_at_)brown(_dot_)name> writes:
Neil> the current infrastructure (of which there are several) are
Neil> fine. But a short term measure that requires lots of people
Neil> to make changes to be really effective is (in my opinion)
Neil> unlikely to be worth the effort.
Authenticating the sender arguably has long term benefits; it's just
that authentictating the sender won't necessarily (in itself) stop
spammers.
Yep. And as long as we are talking about "authenticating" something I
am reasonably comfortable with the discussion. But when we start
talking about "rejecting spam" or even "rejecting mail based on
MARID", I get worried. The issue of what can or cannot be rejected is
very separate from the issue of what can or cannot be authenticated,
and how. One depends on the other, but don't confuse them.
NeilBrown