Code 2, on Thursday, April 26, 2007, at 4:44:07 PM, you wrote: > I didn't try Agava Antispam Servant, but I have tried BayesIt, K9 and > SpamPal. SpamPal was nearly 100% accurate for a long, long time, but > started missing a large proportion of the HTML stock tip spam. I've > never had a false positive with SpamPal. It's a nice > set-it-and-forget-it spam filter.
> K9 was okay, but it seemed to need constant training (at least it did > for the type of spam I get). It also gave a lot of false positives, > so I had to monitor my spam folder regularly. > I never understood why, but I was getting about 10% accuracy with > BayesIt despite training it with hundreds of spam and non-spam > messages. I was frustrated, so I dumped it. > As you can see, my criteria for a good spam filter are accuracy, no > false positives, and ease of training. Free is nice, too :) Thanks. You have saved me a lot of time, and, probably, frustration. Leonard -- Leonard S. Berkowitz [EMAIL PROTECTED] TheBat, version 3.99.3 ________________________________________________ Current version is 3.99 | 'Using TBUDL' information: http://www.silverstones.com/thebat/TBUDLInfo.html