Back in October of 2012 I devised a way of blocking abuse of my mailing list server by detecting mailbait abusers (mailbait.info offers a ‘service’ to fill your, or someone else, mailbox with unwanted mailing list subscriptions). In June of 2013 I refined the technique.
Regardless of the popups, people still try to use mailbait to involuntarily subscribe people to my lists (well, try at least, my lists require a closed loop confirmation system).
Today I received the following threat from a mailbait.info user ..
I’m inclined to ignore the ‘warning’ … but I find it somewhat satisfying that my approach has annoyed someone sufficiently.
As an added measure, I’ve updated my system security mechanism to block any IP that attempts to use the mailbait service more than once.
The problem was, they were a subscriber to the list and had posted before … so the normal counter measures for that didn’t work (the first post for all subscribers are held until approved, to prevent people from subscribing, posting spam, and unsubscribing).
The puzzling thing about this was … the ‘from address’ on the message was not in the subscriber list.
Turns out that Mailman will accept message based on the FROM address of the message or the SENDER address (also known as the envelope-from). The sender addressed is set by the sending mail server and is not normally in the body of the message.
After a bit of digging around, I figured out a way to add this information to the message headers so I can more easily diagnose the problem in the future.
I really wish there was a way to indicate, in DNS, that a domain never sends mail.
That way, if a mail server recieves mail claiming to be from that domain, it can be discarded out of hand.
I’ve got a bunch of domains that JUST do web serving … they never send mail. If the web server that they are hosted on does send mail, it’s sent from via the midrange.com mail server (and is identified as such).