On Fri, 2005-03-18 at 14:46 +0000, David Woodhouse wrote: > On Fri, 2005-03-18 at 10:36 -0400, Patrick O'Callaghan wrote: > > Totally agree (I use SpamAssassin on the server), however Evo > > currently has no way of controlling it. AFAIK, neither does any other > > MUA. > > Some MUAs have a way of graphically editing Sieve filters.
But not Evo. > Some are > capable of feeding mails to sa-learn.That shouldn't be hard, and is > about all you really need, surely? Meaning you have to send the junk message back to the server, but to a special "learn spam" address. Ditto for false positives. You don't gain anything special from Sieve filters in this case; it's still a hack. > I haven't looked into it much because I'm perfectly happy editing > filters in a text editor, and because I don't do per-user Bayesian > filtering. We do. "One man's spam is another man's ham". > It's better to reject mail at SMTP time than to accept it and > then _later_ decide you don't like it, and be left with the choice of > either silently binning it or generating a bounce to a potentially > innocent third party. And if you run SA at SMTP time, the message may > have multiple recipients at multiple domains, some of which aren't > local. So it's hard to do anything per-user, unless you play cunning > tricks with giving 4xx deferrals for any second and subsequent user who > has different SA settings to the first. Our filtering is at delivery time, not SMTP time, since it's per-user. > Running system-wide SA (and other heuristics and greylisting etc) seems > to be perfectly sufficient without anything but the automatically > learned Bayesian filtering, so I haven't really looked hard at getting > feedback from the MUA to SA. I guess it depends on your environment. I can't tell our community (we're a university) "OK, I'm going to decide what incoming mail is spam and you can all take it or leave it". poc _______________________________________________ evolution maillist - [email protected] http://lists.ximian.com/mailman/listinfo/evolution
