Alexander S. Kunz: > But then again, the antispam solution you're using shouldn't catch on > TBUDlist messages... if it does, it would be interesting to see which > parameter raises the spam probability level of a list message - and > correct that error of the spam filter.
SpamCop allows you to whitelist Senders, not Recipients. That's why I set my SpamCop user to pass on the messages as they are received, but with an added header line that inidcates SpamCop/SpamAssassin's findings. Alexander S. Kunz: > [I know why I turned every server side spam filtering OFF - you can never > be sure what it does if you don't have full control over the set of rules] And that's why I created my own system - nowadays I use external anti-spam services nominatively; I don't allow them to decide whether a message is spam, but rather as indicators that "yes, this is very likely a spam". For example - a message that is blocked in SpamCop will be receiving a *very* high spam score (just some points under some preset threshold) - and then, if my system finds chickenpoxes like "v1agr@" in there, the spam score will certainly exceed the Spam Score Threshold - and it's a spam. This system fails very, very rarely. Martin Webster: > I tend to test for lists before generic spam (other than fake messages, > mailing lists etc.). The use of subject tags is therefore redundant even > though many lists to which I subscribe use them. That's how I do it, too. -- St ________________________________________________ Current version is 3.5 | 'Using TBUDL' information: http://www.silverstones.com/thebat/TBUDLInfo.html
