But that still does not help with SpamAssassin. What now?
|Previous Top Next|
Some filters such as SpamAssassin have other tests they use for filtering E-Mail and a score is generated for a particular message. Administrators then use a given score as a "cut-off" when filtering E-Mail spam. Unfortunately, in the process of filtering E-Mail from certain spamware programs, SpamAssassin is also filtering out E-Mail from any Indy-based program (a false positive).
One test SpamAssasin uses is detecting the pattern in the MIME message part boundary which the Indy code has generated. In Indy 10, we have addressed this by making the boundary more random.
In addition, we have made sure that the from header will always contain a name if the name property is blank. We have also forced the In-Reply-To header to always be generated. This helps reduce the score SpamAssassin assigns. We have recently sent E-Mail and received a Spam Assassin score of 0.
Please keep in mind that these aren't the only test SpamAssasin does. There are some other tests involving the content of the E-Mail, some header values, some blacklists such as MAPS, and other filtering technologies. Some legitimate E-Mail messages can generate false-positives with those tests.