> From: Matt Sergeant <[EMAIL PROTECTED]> > > The problems associated with bouncing spam aren't your problems - they > are problems you inflict on the person you bounce to (which almost > always isn't the spammer). > > I get about 10 bounce messages a day from invalid email addresses. I > can't imagine what I would get if spam filters started auto-bouncing > too. All of these are from spam engines which have used my email address. > > Please, do NOT do this. > > If you wish to not receive the spam at all, and let the spammer know > about it, implement a spam filter in your SMTP engine and 550 the > sender. I use qpsmtpd, and it's spamassassin plugin. It's very easy to > extend it to 550 any high scoring spam (but I don't do that because I > want to collect as much as I can ;-) >
That doesn't fundamentally change the problem. In the general case, the message you get might (or for many sites, will definitley) be relayed to you from an upstream provider such as your own ISP. Your ISP's MTA will then try to bounce the message back to the claimed sender, as opposed to rejecting its submission from the actual sender. So, in either case, the claimed sender will get the report. For some sites, it might make a difference, but since the site where I use this (at home, not at work) is using fetchmail to retreive its mail, it wouldn't make any difference at all. The choices are: 1) no blacklisting -- blacklisting has been such an improvement in my fight against spam, that there's no way I'd drop it. 2) manual blacklisting after human review -- doesn't scale, it's just too time consuming. (my home email address is [EMAIL PROTECTED] ... I get way too much spam to wade through all of it by hand) 3) automatic blacklisting without notification -- bad idea, IMO, because there is a likelihood of false positives, so without either human review or notification, automatic blacklisting (or automatic deleting) is just not a good idea. 4) automatic blacklisting with some form of notification (the mechanism I've described, or the slight improvement you describe) -- removes the false positive problem by allowing the sender to appeal. I'm not saying #4 is perfect and without flaws, I'm just saying it's better than the alternatives. John ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
