Jeff Chan wrote:
> It depends on the rate of FPs.  If they are relatively rare, and
> it seems they are, then delists may be an appropriate or at least
> workable solution.
I'd agree it's workable, but I think we can do better. And without
hurting the spam hit rates.

Right now SA's rule structure for this set is laid out in a way that
maximize the spam score while limiting the scores to control FPs.

I'd rather see it structured more along the lines of getting the most
messages on the right side of the 5.0 line.

>> Can we address the real question here:
>>     
>
>   
>> How can we keep the spam tagged, and try to mitigate the FPs by keeping
>> additive scores for multiple URIBLs more moderate? +20 worth of URIBL
>> hits is fine on spam, but astronomically high scores don't really help
>> SA when the tagging threshold is +5. However, they do hurt SA when
>> overlapping mistakes happen.
>>     
>
> Perhaps that's an appropriate question for the SpamAssassin
> developers.

Yes.. which is exactly who I was primarily trying to reach by posting
here on the spamassassin, before this turned into a large
misunderstanding between the URIBL operators and myself.

I posted on the users list instead of the dev list because I wanted to
invite commentary from the URIBL operators and other users. Me and Chris
S were actually having some good discussions along these lines.
Unfortunately, along the way I touched a few nerves.

Reply via email to