>> I want to understand EXACTLY how my mail is triaged and why, >> particularly on my critical accounts.
MA> Unfortunately, it is my sad yet considered opinion the choice to do MA> your own filtering is evaporating. My friend told me boastfully about MA> how his Iowa (USA) ISP was "Filtering" his mail with a bayesian MA> filter. I pointed him to a free bayesian filter he could operate MA> locally, knowing exactly what was getting bounced. I politely beg to disagree. I think simple self filtering of spam is easy for most people. More than 95% of all my wanted email is both addressed to me and comes from someone with whom I have previously corresponded in my address book (2000 names), or comes from a dozen or so servers (eg. my University). Of the remaining 5% of wanted mail, it is difficult to imagine anything that would fail to mention my name (Dear Mark, Hi Mark, Mark, Hello Mark) or a few dozen keywords that any novel new person approaching me would have to mention to be of any interest whatsoever. Clearly this doesn't apply to everyone who uses email, but I would guess it applies to 95% of us out there who use our email addresses with a modicum of discretion. A few simple filters help to ice the cake (not addressed to me, multiple similar addressees, a few nasty keywords, foreign characters in subject, and routings through a few countries through which legitimate mail to me would never be sent). I really think it is OTT to suggest that The BAT! alone couldn't suffice for most people. The critical aspect of spam detection is *NEVER* to miss *IMPORTANT* legitimate mail and I think it is unlikely I would. Letting a few spams through is hardly relevant. ________________________________________________ Current version is 1.62r | "Using TBUDL" information: http://www.silverstones.com/thebat/TBUDLInfo.html
