https://issues.apache.org/SpamAssassin/show_bug.cgi?id=5503
--- Comment #4 from Kevin A. McGrail <[email protected]> --- (In reply to comment #3) > (In reply to comment #1) > > feature request still stands and would make it simpler that "easy enough to > circumvent that if you want to" For the record, I'm not discounting the feature request. I just lowered the severity to indicate that it was not severe enough to warrant blocking a release. Critical + Blocker do warrant a blocker per our project's conventions. Looking at this issue, I noted two things: First, based on AutoLearnThreshold.pm, Note that certain tests are ignored when determining whether a message should be trained upon: * rules with tflags set to 'learn' (the Bayesian rules) ... * rules with tflags set to 'noautolearn' So the first note is that there is no current functionality that achieves this goal and the docs if they lead you to believe it exists are wrong. Second, it appears the logic in "lib/Mail/SpamAssassin/Plugin/AutoLearnThreshold.pm" only checks for the 3 body +3 header requirement for spams which makes sense. That makes sense but I wanted to clarify it for the purpose of this discussion. Give me a few minutes and I think I have working code for this change. -- You are receiving this mail because: You are the assignee for the bug.
