It's been a long time since many spamming techniques are around and that is why we need spam filtering and prevention software. But is there a realistic chance that sometimes we will have no spam in our mail boxes? In my opinion, this is not possible. At least to a certain level of possibility. The best available anti spam solutions come probably from open source community and thus are open to anyone. To the spammers as well. And here we have the source of the problem. The spammers will always have the advantage of knowing more. They can simply analyse how all the anti spam solutions are working and can simply override the filters by e.g. including some improbable words in their e-mail messages. Spam filtering software must always have some weak points - for instance - there must not be any 'false positives' - means that it is absolutely unacceptable that a spam filter would filter out a 'real' message. To avoid this, there are some rules in each spam filter to be applied - mainly based on statistical approaches or so. This can however be outscored by spammers really easily. The amount of unsolicited content can be lowered to as little as approximately 95%, but cannot be eliminated completely. I think that the best solution available today is a combination of a server side spam filtering software with a cooperating client side software. The server part should take care of a basic bayesian filtering with regular automatic server spam database updates, and the client software - which should primarily reside in the e-mai client software (so called MUA - mail user agent) should automatically maintain a database of e-mail addresses the user sends responses to and consider such addresses as 100% negatives. This can prevent almost all false positives. In the near future, however, there seems not to be any chance of developing a really effective anti spam solution which would not let in a spam meesage and make 0% false positives.