On 01 Mar 2019, at 07:21, Thomas Seilund <t...@netmaster.dk> wrote:
-- Once a day for each user I clear the bayes files and rebuild bayes files
with:
On 02/03/2019 13.38, @lbutlr wrote:
You are removing the bases entries daily and rebuilding them based on a very
few (if any) messages in your LaernAs folders?
That’s the same as not using bayes at all.
On 03.03.19 11:27, Thomas Seilund wrote:
Each user has a ham mail folder and a spam mail folder.
I instruct user to have at least 10 not spam mails in the ham folder.
spamassassin needs at least 100 pieces of each to start hitting.
And I instruct the users to move spam that make it to the inbox to the
spam folder.
In most cases users have +10 mails in the ham folder and +100 mails in
the spam folder.
How should SA learn from the two folders if not by running sa-learn on
each of the two folders regularly?
note that the commands you have mentioned:
https://marc.info/?l=postfix-users&m=155145015801188&w=2
don't clear bayes database, they only train new spam/ham.
the complaint was about removing bayes database daily. You don't need to
remove bayes database at all. Don't clear the bayes database.
I use SA by integrating SpamAssassin into Postfix using spamd and I
wrote a bash script to rewrite spam method
as described in
https://wiki.apache.org/spamassassin/IntegratedSpamdInPostfix
there are better way to integrate spamassassin to postfix, I'd recommend
spamass-milter if you weant to keep per-user databases.
--
Matus UHLAR - fantomas, uh...@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
Honk if you love peace and quiet.
