From: Gary V <mr8...@gm...> - 2008-03-27 20:14:12
|
On 3/27/08, Mark Martinec wrote: > Gary, > > > But what if you have one recipient in a @bypass_spam_checks_maps and > > another that is not? Would you still not need to place the recipient > > in the @bypass_spam_checks_maps in the @spam_lovers_maps? > > @bypass_spam_checks_maps is pretty much independent from everything else. > If at least one recipient requires spam checking (has bypass at false), > then spam checking is performed. Or in other words, if all recipients > have bypass at true, spam checking need not be performed (but the result > may still be available, e.g. when cached). > > What happens next depends on whether mail is considered spam (above kill level > for this recipient), or not. If it is not spam, it is delivered normally > and not quarantined. If it is spam, it gets quarantined unless quarantining > is disabled, and it is blocked, unless recipient is a spam lover. > > As you can see, if mail is not considered spam (e.g. when mail is clean > or recipient has a high kill level, same thing), the @spam_lovers_maps is > irrelevant. > > See first couple of items under: > http://www.ijs.si/software/amavisd/amavisd-new-docs.html > > Mark > I see. Oops, for some reason I must have casually assumed that because mail is tagged as spam at tag2_level, this alone might be sufficient to short circuit delivery of a message to a recipient in @bypass_spam_checks_maps. Now I imagine this is actually a result of $final_destiny being set to anything other than D_PASS. Thus, if kill_level is beyond any possible spam score, the action in $final_destiny never takes place. All spam is delivered. -- Gary V |