Landmark wrote:

> > So, you're saying that without an MX record, some spammers
> > do/did not bother to send their trash (to you) at all?
>
> No,

Yes.

> he's saying that with his misconfigured server he believes his
> spam has dropped off

"Belives my spam has dropped" ?

How about "I know my spam has dropped" you fool. Or don't you believe
that you can count the number of delivered spam to a given account?
I've posted my numbers before, and you have nothing to say each time I
do.

> which may well be true in the case of "fire and forget" spam

"fire and forget" spam? When did you invent that term?

The more commonly used term is "zombie spam" or "direct-to-mx" spam,
which is by far the most prevalent source of e-mail spam for the past
5 years.

> but more importantly, he beleives no legit mails are
> being afffected, but he has never tested this,

And how does one go about testing this? Do you propose asking the
postmaster of every domain on the planet to send you an e-mail?

Our organization has been in business for 15 years. We've been
operating various e-mail accounts at our domain since 1996. We are in
daily e-mail contact with a base of a few thousand clients in Europe,
Asia, Australia, US/Canada. We have experienced no interruption in
e-mail communication with any of them over the past 2 years, and
during those 2 years we engaged in e-mail contact with a few hundred
new people. That's proof enough for me that A-record fallback when MX
lookup fails is pretty much universal (and reliable) behavior.

You can be argumentative and bull-headed and say it's not proof
enough, even though it's in the RFC's, and even though there is no
practical test that would satisfy you anyways.

> only asserted that since he is getting some mail he therefore
> believes he is getting all mail. There is an obvious flaw in
> his logic.

Given the base of prospects from which we would expect to get e-mail
from, and given the 2-year span that we've been operating without an
MX record, then according to your logic we should have received at
least one communication from someone (either by phone, or from an
alternate e-mail account, or by fax) that they are having problems
e-mailing us. We have received no such communication.

> This is in marked contrast to TGP who is doing a proper test
> of NoListing. The author of NoListing is also markedly different
> from SpamGuy in that he carefully sets out the limitations of
> the method, and the things to take into account before
> adopting the NoLisitng technique.

The author of NoListing is proposing an MX configuration that could
also result in the delivery failure of legit e-mail from MTA's that
don't know how to properly handle serialized servers combined with the
strategic failure of some of them. In fact, I'd argue that such a
configuration is more prone to legit-delivery-failure vs a simpler
mx-lookup-failure strategy.