On Wed, 26 Feb 2003 11:33:00 GMT Angel Rivera <[EMAIL PROTECTED]> wrote:J C Lawrence writes:
Which are precisely the reasons I don't and won't do SPAM filtering at the MTA. The potential cost of error is high, and almost all ability to supervise and correct is lost.
A lot of these problems are probably due to misconfigured tools.
Accepted, partially. Various RBLs have a tendency to mark associated netblocks, which I find deceptive.
ah, but I don't find those deceptive-at least in those I use. SPEWS, for example will expand the blocked IPs if they do not act on complaints-the theory being to give them a little finacial push for them to do the right thing.
I am a firm believer of RBLs and we use that as the first line of defense against spam. If it has come to the point where someone is on one of the RBLs we use-we need a break.
To that we have added SpamAssassin in tagging mode. The few false positives that it catches are simply tagged and can be whitelisted.
I use RBLs, SpamAssasin, Razor, TMDA and a few privately developed filters in concert at LDA time to assist in correct folder filing.
I do use razor and have submitted some spam. I don't use TMDA-more nuisance than I want to deal with. I do like the idea of configuring SpamAssassin so each user can configure their own whitelists as they choose, so I will most likely be heading in that direction.
-ar
