Thank you Thomas,
that is exactly what I have been doing for years and it worked fine. 
MaxBayesFileAge was the only thing I adjusted to keep ham and spam collections 
at a litte more than 14000 mails. Not more to avoid wasting disk space 
unnecessarily.
Until recently I have been dealing with very low mail volume, leading me to 
MaxBayesFileAge of say 750 days for spam and 250 days for notspam.
Now the mail volume has increased much and - continuing the same strategy - i 
reached a current ratio of 650 days for spam and 24 days for notspam.
At that point I wondered if it is healthy to build a corpus of only young ham 
and young to very old spam.

I am also looking for other ways to slow down notspam collection, e.g. content 
based by excluding automatic status reports and such.
Currently this is all academic - ASSP works absolutely fine for me. :-)

Regards
Dirk

------ Originalnachricht ------
Von "Thomas Eckardt" 
<[email protected]<mailto:[email protected]>>
An "Dirk Kulmsee" <[email protected]<mailto:[email protected]>>; "For 
Users of ASSP" 
<[email protected]<mailto:[email protected]>>
Datum 26.08.2026 15:14:15
Betreff Re: [Assp-user] Question about freqSpam / freqNonSpam

keep in mind, that notspam mails are required to maintain the spamDB and HMMdb 
(e.g. Bayesian and HMM)!

I would leave ' freqNonSpam' at 1.
If the notspam collection grows too fast and you use Bayesian or HMM, change 
'MaxBayesFileAge' to limit the amount of collected files.
The rebuild task processes youngest files first (down to the oldest). This way 
both databases are based on the most recent processed mails.


Thomas

DISCLAIMER:
*******************************************************
This email and any files transmitted with it may be confidential, legally 
privileged and protected in law and are intended solely for the use of the
individual to whom it is addressed.
This email was multiple times scanned for viruses. There should be no known 
virus in this email!
*******************************************************



Von:        "Dirk Kulmsee via Assp-user" <[email protected]>
An:        "[email protected]" <[email protected]>
Kopie:        "Dirk Kulmsee" <[email protected]>
Datum:        26.08.2026 10:49
Betreff:        [Assp-user] Question about freqSpam / freqNonSpam
________________________________

[cid:_1_11B8A0AC11B89CC40048B7A2C1258E5F]       
[cid:_1_11B8A9E011B8A5F80048B7A2C1258E5F]




Hi everybody,
I have a question about the frequencies for collection. The description says I 
have to collect every mail or else BlockReports cannot recover mails if they 
haven't been collected before. That sounds obvious. But how about notspam 
mails? If ASSP considers a mail as good, it is delivered to the recipient and 
there should be no reason to resend it by a BlockReport.
Question: can I safely lower the Non Spam Collection Frequency (freqNonSpam) 
without losing important data or am I missing something?

Dirk_______________________________________________
Assp-user mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/assp-user


_______________________________________________
Assp-user mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/assp-user

Reply via email to