Hello Fabio, This would seem to be the two lines of interest here:
FC> 05.11.2004 20:19:42 Non-spam frequency dictionary: FC> 05.11.2004 20:19:42 Error loading file. May be it is corrupted or lost. It sounds like the file that holds the "non-spam" info us corrupted (it might also be missing, but that seems less likely). As far as I know, the only real solution is to delete the offending database files, in this case, the "non-spam" (or ham) data. Look in your data tree for the directory '/BatMail/bayesit/base' for three files that start 'nspamdict' There is a '.bye', '.idx' and a '.lst'. Delete those. Next you will need to find some "good mail" to begin re-building up your "ham" database or everything will end up looking like spam. Because I try a lot of beta versions I keep a good number of both spam and ham messages so I can retrain when things "go wrong". But if you don't have a lot of "ham" messages sitting around, you might be better served if you start your training over from scratch. You do that by deleting all the files in that directory and just start again from scratch with the training. It's a tough call as to which would be the better choice. It depends on how much mail you have sitting around for that training. Or maybe I am just too paranoid about "false positives" in my spam filtering <g> -- Best regards, MikeD mailto:[EMAIL PROTECTED] Using The Bat! v3.0.2.1 on Windows ME 4.90 Build 3000 ________________________________________________________ Current beta is 3.0.2.6 Rush | 'Using TBBETA' information: http://www.silverstones.com/thebat/TBUDLInfo.html IMPORTANT: To register as a Beta tester, use this link first - http://www.ritlabs.com/en/partners/testers/
