Am 26.02.2016 um 15:15 schrieb RW:
On Fri, 26 Feb 2016 14:30:23 +0100
Reindl Harald wrote:

score VERY_LONG_REPTO_SHORT_MSG             3.999 3.999 3.999 3.999
header    __VERY_LONG_REPTO             Reply-To =~ /[^\s\@]{20,}\@/

Reply-To: malsorzata.warmin...@oranet.pl

very long?
20 chars?
4 points?
seriously?

that needs to be lower scored or 20 raised to much higher values

or perhaps include ".+-_"  in the list of excluded
characters - it's pretty reckless as it stands

$ printf "<Richard.Milhous.Nixon" | wc -c
       22

$ printf "<homer.simpson+amazon" | wc -c
       21

even the SHORT_MSG part is questionable, the FP was a hotel booking request and we scored that down to fixed 0.5 points

* 9 not very long lines of normal content
* --
* --
* 8 signature lines, each prefixed with ":: "

combined with
 * score REPLYTO_WITHOUT_TO_CC 2.399 1.946 0.607 1.552
 * score MISSING_HEADERS 0.915 1.207 1.204 1.021
 * score BAYES_50  0  0  2.0    0.8

the rule above is a posion pill, hits here 50% ham and 50 % spam while the spam would have been rejected anyways

for 12 hits in the complete month not worth the troubles of a FP

Attachment: signature.asc
Description: OpenPGP digital signature

Reply via email to