[Spambayes] feature request

Tony Meyer tameyer at ihug.co.nz
Tue Aug 10 08:24:49 CEST 2004

> Just to be a devil's advocate.....


> 1. Many (most?) user communities will not accept a 
> client-based solution to the spam problem. Many users are so 
> spam-averse that merely seeing it, even after it has been 
> identified and collected, makes them go ballistic. The more 
> Spambayes can be made to resemble a server-based system, the 
> more widely it will be accepted. I wouldn't use the proposed 
> feature myself, but there it is.

As long as the person willing to put the code in is also willing to deal
with the people that then lose mail they want because SpamBayes has moved it
to Deleted Items by mistake (i.e. a false positive), then I'm fine with it.
I'm not such a person.

Besides, Outlook users can just use the auto-archive feature, and
non-Outlook users can already do this.  Seems rather pointless.

> 2. I'm under the impression that a few months ago someone 
> (maybe more than one someone) proved that if you start with 
> an empty database, Spambayes learns quickly, and that within 
> a few days it performs almost as well as it would if trained 
> in the customary way. If so, saving spam for retraining 
> should be a non-issue.

Yes the text that says that (the FAQ, from memory) is a bit out of date.
"Proved" is a bit strong, but generally keeping mail around for retraining
isn't necessary.

=Tony Meyer

Please always include the list (spambayes at python.org) in your replies
(reply-all), and please don't send me personal mail about SpamBayes. This
way, you get everyone's help, and avoid a lack of replies when I'm busy.

More information about the Spambayes mailing list