Hi, Hopefully I didn't miss this one in the archives...
I've been playing around with qmailadmin (1.0.23), specifically, the "enable spam blocking" option. It works great, and I've got a nice maildrop filter built that also allows for each user to have their own filter rules. The only problem I see is that if there is a catchall account and you enable spam-filtering, none of the catchall mail is filtered. The reason being is that the .qmail-default for the domain effectively bypasses the user's .qmail file, and therefor, the filtering. How can we work around this? Would it be best to have qmailadmin quietly stick a .qmail-default file in the domain directory that calls another maildrop filter? That's what I'm doing manually now; I have two sitewide filters, "mailfilter.user" and "mailfilter.domain". The user filter is a simple set of rules that passes the message through spamc (if it's under X bytes), and either delivers it to the inbox if it's clean or if it's dirty, it checks for a "Spam" mailbox (and creates it if it doesn't exist) and delivers it there. The domain filter does the same thing, but hands the message over to vdelivermail. Would it be worthwhile having qmailadmin put a filter in place on the domain if the catchall user is set to have spam filtering enabled? We could include some maildrop samples and directions in the distribution... Thanks, Charles
