My bad, I was thinking of the Mercury plugin version which is what I typically use (I used it long before the Pegasus version was added so I've stuck with it).  I'm surprised the Pmail version doesn't include this feature...

 

First of all I want to say that Spamhalter is a fantastic tool to separate spam from ham. [Y]

I highly trust Spamhalter and would appreciate an additional option to delete spam e-mails automatically (or to move them into the Pegasus recycle bin) if reaching a certain spam probability level.
That doesn't mean that I want to delete all spam e-mails automatically (by the way: I am speaking about Pegasus Mail, not about Mercury).

I'm looking for an easy way to get rid of e-mails that have a spam probability of 99% or more (for example).

Cheers! 

Marc

 

Well a method might be to make the deleted messages folder the target for Spamhalter (NOT TESTED!)

[quote user="Han v.d. Bogaerde"]

Well a method might be to make the deleted messages folder the target for Spamhalter (NOT TESTED!)
[/quote]

 
Your suggestion means that all spam messages are moved into the deleted messages folder. That's the way Spamhalter works at the moment.

Spamhalter is capable to separate spam from ham, but after that distinction there is no possibility to differentiate between clear-cut spam (99% or 100% spam probability) and "unclear" spam (e.g. 80% spam probability).

Marc
 

This is easily accomplished using custom filtering rules.  Just regex match against the Spamhalter headers.

[quote user="Barius"]This is easily accomplished using custom filtering rules. Just regex match against the Spamhalter headers.
[/quote]

Hm, I don't know what you mean by "Spamhalter headers". AFAIK there is no connection between the Spamhalter results and the possibility to use custom filtering rules.

Marc
 

Hello!

I also think that there are no Spamhalter entries in the raw view you could evaluate for filtering rules. As far as I understand Spamhalter, there is a simple "yes"-"no"-classifying: a message is either good (no spam) or evil (spam), but there is no intermediate classification.
You may want to use an external spam plassification program. POPFile is one of them, also able to classify incoming messages in "buckets". As usual, setting up a Bayes filter tool needs some training (and in case of POPFile: the mroe buckets you have, the moren training it needs); and an external program needs to have a link between itself and Pegasus Mail (i.e. you have to edit Pegasus Mail's POP3-settings, as far as I know).
In short: this would mean some extra work, so you have to think about whether your goal justifies that.

A German-speaking manual for POPFile can be found here: http://popfile.sourceforge.net/manual/de/manual.html.