I also realize that it has probably already been discussed in the
appropriate places, but since I'm not following those discussions I
offer it here for consideration.
It seems to me that to answer Chris' legitimate concern about content,
a different approach, not limited to FAQs, might be incorporated,
consisting of two parts:
1) tighten up the meaning of "approved" so that the approval header
amounts to a crypto signature of the article by the
moderator. This would dramatically reduce the likelyhood of
forged approvals.
2) exempt moderated groups from spam BI calculation, other than
checking the signature.
I realize that I'm suggesting a significant change in the software, as
well as a process that might be CPU intensive. However, it might not
be necessary for the signature to be checked by most sites, only by
spam-cancellers.
marty
-- marty fouts mailto:fouts@null.net http://www.best.com/~mjfI believe in censorship. After all, I made a fortune out of it. -- Mae West