Question about Mutt macro and output

[David Relson]

Hello Daniel,

I can't answer the mutt questions because I don't use it.  The other 
questions I can and will answer.

At 09:26 AM 3/19/03, daniel wrote:

>One last question: in the given .procmailrc examples:
>
>        :0HB:
>        * ? bogofilter
>        Mail/spam
>
>        :0HB:
>        * ? bogofilter -u
>        Mail/spam

Both the above lines classify the message and divert it (if 
appropriate).  You only need 1 of them.

>and
>
>        :0fw
>        | bogofilter -u -e -p
>
>        :0e
>        { EXITCODE=75 HOST }

Exitcode==75 tests for X_TEMPFAIL (see sysexits.h).  The test is useful and 
should always be in your procmail recipe


>        :0:
>        * ^X-Bogosity: Yes, tests=bogofilter
>        spam-bogofilter
>
>
>am I correct in assuming that we only need to use one set of these 
>commands (I am using the second set)?

Yes.  The first command "set" diverts spam using bogofilter's return 
code.  The second set diverts based on content of the X-Bogosity line.

>Could anyone supply some more useful macro commands for Mutt so that we 
>can see other interesting statistics and information about how the filter 
>is working?
>
>One last question:  I have just set up bogofilter recently.  Here are the 
>counts so far:
>                        spam   good
>.MSG_COUNT               92    234
>
>about how long does it take before it will start actually recognizing and 
>filtering spam?

Bogofilter will classify messages even with a total MSG_COUNT of 
1.  However, for it to do a good job, tokens in the incoming messages need 
to be found in the ham and spam wordlists.  How soon bogofilter starts 
being useful depends on how many words are in the wordlists.  The number of 
messages is less important than how many words bogofilter recognizes.  The 
following command will show you the word counts:

for file in $BOGOFILTER_DIR/*.db ; do echo $file; bogoutil -w $file | wc -l 
; done