Mike Holling writes:
> I just fear that it's a very small step to go from blocking known dialup
> pools to blocking any IP that resolves to a pattern like
> "1-2-3-4.example.net". That looks like a dialup, and if it's a cablemodem
> or DSL line who cares, that may as well be a dialup right? After all, if
> it were a true mail server that could be trusted it wouldn't have a name
> that starts with its IP address, only dialups use those.
Well, that's life. My personal policy is simple. On the receiving side, I
implement mechanisms to block dialup spam in a manner that I believe is far
more efficient than outright blacklists. Others have reported that
filtering out dialup spam based on the headers, instead of IP addresses,
yields too many false positives, however my experience doesn't show that.
As far as I know, I've never blocked a non-spam message sent to my
spam-filtered address from a dialup, and have blocked all dialup spam,
sometimes sent from the same IP address pool. I have become aware of some
isolated cases where the spamware won't be blocked by my filters, but so
far I have not been on the receiving end for now, and I've decided to cross
that bridge when I come to it.
On the sending side, for a very, very short list of receipients who I have
a specific reason to regularly send mail to, and who I know are blocking
mail from dialups, I make sure to smarthost my mail to that domain. In all
other cases, if my reply to an inquiry bounces due to an obvious dialup
block, I will manually resend that reply, but all further E-mail from the
same domain will be blocked as well, with an informative rejection message,
inviting the sender to try again from a different domain.
I no longer get excited over this issue, and prefer to abstain from
religious flame wars over this. That's simply how I'm handling it, and, so
far, the approach works rather well.
