John Levine <johnl(_at_)iecc(_dot_)com> wrote:
The decision about whether it's worth the cost of redefining the
message bounce address as a sender address and adding IP based
verification goop for that address is a political one, not a
technical one.
If that's true, then I'm much more skeptical that the MARID will be
deployed. Asking people to upgrade their systems "because it works,
trust us" is not good practice. If the benefits are quantifiable, on
the other hand, the decision becomes much more technical, and less
political.
Could you explain in detail how you're prevented from implementing
LMAP?
Not me, specifically. But I see a lot of FUD around the topic.
"LMAP will cause the sky to fall!". Such FUD cannot honestly be said
to be supporting LMAP, encouraging people to deploy it, or dealing
with it's benefits and drawbacks in a professional manner. As a
result, it looks like the purpose of the FUD is to try to prevent
people from implementing LMAP, based on non-technical reasons.
Or maybe the FUD is there for constructive purposes. Who knows.
And to ask my perennial idiotic question: How does the recipient
tell that the message is valid?
The same way they have for the past 20 years, I suppose. How do you
tell that messages from this list are valid? If you can't, how can
you bear to accept mail from it?
For systems with low spam and abuse rates, it's trivial to tell.
For systems with high spam and abuse rates, it's a lot more
difficult. Enter various proposals.
The problem I see is that the only difference between those two
systems is luck. Some are lucky enough that abuse isn't a problem,
and others aren't. So there's a natural division between the two,
into groups which do see a problem, and ones which don't.
Alan DeKok.