For example in the above, if it finds anything in the body that indicats it came from the MSN Communities, it gives it a score of 0.0 (which if I am correct doesn't add anything to the total spam test score), whereas on the test marked: Q_FOR_SELLER, if the Subject is an eBay question, then depending upon the type of scoring system you use it would give it a score of any the above listed.
Correct, and to be more specific the choice of which score is based on what additional features you have enabled.
If you run with rules-only it uses the first score in the list
If you run with rules and network checks (ie: dnsbl's, razor, etc) it uses the second score
If you run with rules and bayes it uses the third score.
If you run with rules, network checks, and bays, it uses the fourth score
Each of the scores is tuned by the GA (Genetic Algorithm) to match up with those conditions.
This is the problem, what is the minimum setting you can use? what is the maximum
I'm not quite sure what you mean by "minimum setting".. individual rule scores can be just about anything, but for practical purposes most rules wind up being assigned scores by the GA that are between -5 and +5. Sometimes it goes a little over that range, but I've never seen it go as far as +10 or -10 in a final scoreset.
A few "sure fire" rules, like blacklist_from, are hand-set to very high scores like +100, but the vast majority of the scores are assigned by the GA.
how do they increment?
What do you mean by "increment" here? If you're wondering how it chooses which of the 4 scores to use, I explained that above. If you are wondering how a message which matches multiple rules is score, SA just adds the scores together.
as some are makred as -0.443, -0.536, then some are marked as 4.300, 0.218
Whilst I understand the total spam score is usually set to 4.0 (this is different than this total score, basically from what I understand, if any of these tests add up to 4.0, then they are tagged as spam.
Actually, the default spam threshold is 5.0, not 4.0, but it can be reconfigured.
Also mail only gets tagged as spam if all of them add up to over the threshold, not if any of them add up to over the threshold.
So a message that has 4 rules match scoring 4.0, 1.5, -2.0, and -1.0 would wind up with a score of 2.5. If your threshold is over 2.5, the message would not get tagged, even though the first two rules add to 5.5.
Sorry if I do not make myself clear, hopefully I've come across as asking the correct quuestion.
Well, I'm trying, but your choice of words is pretty confusing here.
------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
