About > News > Stigmergic antispam > Discussion


On the pros and cons of this antispam technique see also spamtrap page on wikipedia.
  • backscatter generates false positives: Other MTAs could generate service messages (bounce, undeliverable mail), which are not considered spam, if spammers forge a fake sender address using one of our fake addresses. These service message, if directed to fake address, will blacklist regular MTAs and not spammers. See how qmail and postfix cope with backscatter.
  • mailing lists generate false positives: if a spammer registers a fake address in a mailing list where other users from our sistem are registered, the first delivery to a fake address will deny further deliveries to regular addresses.
  • probabilistic approach: since spammers deliver mail in large lots, the probability for our users to receive spam depends on the prevalence of fake address on regular ones. Spam will be delivered until the first fake address is used.
  • handcrafted spam addresses lists: if spammers spot the fake addresses, they can clean up their lists and the technique is useless. But this should take them busy for a long time, make the cost of sending spam higher and improve the quality of spam eventually delivered :-). Changing the list of fake addresses from time to time should help.