Just to chime in a note on this topic. The default scores (as distributed by upstream) and dynamically generated by having the rules analyze corpuses on known spam and known ham. Based on which rules match on the SPAM and on the HAM the scores are computed in a way to minimize the number of false positives and false negatives. This means that if a score ends up with a high positive score that it is a good indicator for SPAM (at least based on the corpus it was run against). The opposite is true about large negative scores and HAM.
For more information check out http://wiki.apache.org/spamassassin/HowScoresAreAssigned ---------------- Thanks Jefferson Cowart [EMAIL PROTECTED] > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] > Sent: Sunday, January 16, 2005 11:47 > To: Michelle Konzack > Cc: [EMAIL PROTECTED] > Subject: Bug#290520: spamassassin: please do not criminalise > end users, users behind dynamic IP, lower or remove DUL test > default score > > Michelle Konzack <[EMAIL PROTECTED]> tapota : > > > Note: I am not Maintainer of this package. > > > > Am 2005-01-14 16:54:08, schrieb Mathieu Roy: > >> Package: spamassassin > >> Version: 3.0.2-1 > >> Severity: normal > > > >> 0.1 RCVD_IN_SORBS_DUL RBL: Envoyé directement depuis > une adresse IP dynamique > >> [151.24.72.136 listed in > dnsbl.sorbs.net] > >> 1.7 RCVD_IN_NJABL_DUL RBL: NJABL: Envoyé depuis une > adresse IP dynamique > >> [151.24.72.136 listed in > combined.njabl.org] > >> > >> > >> While it was effectively a spam caught, DUL/Dynamic IP > scoring (1.7) > >> is quite problematic. > > > > Why ? - Lowering let around 50% of the SPAM in > > Not in my experience. They are others effective tests. More effective > as they rely on something else than the connectivity type of a user. > > > >> - The score increase with each DNSbl you use, even if > they use the > >> exact same source > > > > Thats right. > > And that would make sense to you? > > > > >> - Why the hell freedom to run a server should not be > given to someone > >> behind dynamic IP? Restricting users freedom is a very > sensitive issue, > >> I do not think software shipped by Debian by default should > >> criminalize dynamic IP users by assuming they are > guilty of something. > > > > This is a protection for innocent $USER. > > I do not allow any Messages comeing from DUL. > > On what basis? Who the hell ever decided that people behind dynamic ip > are not entitled to run an smtp server? > > What you call "protection" already harmed "innocent $USER" on some > mailing list run on server I administrated, because we realized to > late that these DUL tests could raison the spam level indicator of > mail high enough to get these mails rejected. > > I do not call that protection, I call that dumb assumptions, blatant > ignorance of what users of the internet are allowed to do legally. > > > > >> So, would it be possible to have the DUL test deactivated > by default, > >> or with a score less important (0.1 is fine)? It is good > that a piece > > > > Do you know, that you can lower the SCORE from > > > > ___( '/home/michelle.konzack/.spamassassin/user_prefs' > )______________ > > / > > | score RCVD_IN_SORBS_DUL 0.1 > > | score RCVD_IN_NJABL_DUL 0.1 > > > \_____________________________________________________________ > _________ > > > > Sure, the problem at stake here is the fact that the default settings > are so high. > > If you are happy with such tests, you should be the one raising their > default score. Because these tests have no valid ground, unlike razor, > pyzor, bayes, regular DNSbl tests (open relay lists, known spammers). > > > >> of software like spamassassin get distributed widely. Spam will be > >> less and less commercially interesting. But it should not cost end > >> users freedom to run a server, don't you agree? > > > > If spamassassin lower the score, we get two times more SPAM in our > > boxes as with this high score. I do not like to chnage some > 100 Boxes > > to higher score. 1.7 is right. > > I talk about freedom to run an smtp server, you talk about your > experience of spamassassin. Off-topic. > > > > And, - there are very less $USER, which send messages directly, > > but millions of SPAMesn SPAMbots and Viruses. > > That's a point of view. It is also off-topic. > > > >> People that run DUL DNSbl warn users about DUL list usage. > The problem > >> here is that someone may filter DUL users without even noticing it, > >> just like if it were an absolutely reliable anti-spam criterion. > > > > I do not like DUL-Sender, because never had someone which had send > > > a Messages directly and I am working since many years with Linux. > > Huu, you already received two! My messages are sent directly, even if > I'm not behind a dynamic ip. > > > > > -- > Mathieu Roy > > > +------------------------------------------------------------- > --------+ > | General Homepage: http://yeupou.coleumes.org/ > | > | Computing Homepage: http://alberich.coleumes.org/ > | > | Not a native english speaker: > | > | > http://stock.coleumes.org/doc.php?i=/misc-files/flawed-english | > > +------------------------------------------------------------- > --------+ > > >
