Re: [Evolution] How to handle unsolicited e-mails with evolution 2.32.2 ?

[Thomas Prost <thomas prost prosts info>]

Am Sonntag, den 27.11.2011, 13:10 +0000 schrieb Pete Biggs:
> Well it's fairly obvious? You've been going on for ages about it not
> filtering your messages, therefore I thought it was fairly safe to
> assume that, err, it wasn't working as intended.
YouÂre definitely right - if my intention on using bogofilter was right.
Sadly seems to me, all documentations IÂve found "started in the
middle". I couldnÂt find any lines about the basics of bogofilter
(otherwise IÂd known, who this Mr. Fisher is ;-)

>  bogofilter -t message.file
>
> this will print 'Y' if it's spam, 'N' if it's not and 'U' if it's
> unsure.  There are other command line options, do 'man bogofilter' to
> find them.  It's not difficult finding out this information.
ThatÂs approximately what I was searching for, but not a single mail for
a test. That wouldnÂt say anything about the confidence of my data.
I was looking for kind of test data set with a known desired output, so
that I can analyze, how good my data set is already "trained".
Only seeing in alldays running, that the spam-folder still remains the
same - empty, is very disappointing without knowing why ...
(as Ross Vandergrift wanted to point out, too, I assume ...)

--
getting on,
Thomas