Basically I believe spam cancellors can leave *.answers alone so no
BI index needs to be considered if any "*.answers" group is found in
Newsgroups: list. However, this means the moderator team needs to
handle all forged approvals in "*.answers", which might not be an
easy job because each FAQ received at MIT needs to be compared with
a database to check if the approval is forged.
Two problems could go with this concept:
(a) A deliberate spammer could add a *.answers into his Newsgroups:
list, forges an approval, and makes the MIT server "blind" for
this spam. This could be solved by an unpublicized site
monitoring *.answers activity.
(b) A vicious person could forge an approved FAQ by a real FAQ
maintainer and then spams the scandal/libel post to many
groups. I cannot think of a good solution except some
encryption works.
Any comments about this thought?
A few weeks ago I showed interests about participating in the
moderator team to relieve the load, but later my busy schedule
stopped me from participating at least temporarily. I hope I
would not increase Pam's load.
=======================================
Forwarded message:
> Date: Tue, 10 Mar 1998 14:11:35 -0500 (EST)
> Message-Id: <9803101911.AA19033@moe.optics.rochester.edu>
> From: Pamela Greene <pgreene@optics.rochester.edu>
> Subject: Re: permitted cross-posting for genuine FAQs
>
> Uri Raz wrote:
>
> > How about simply not counting any of the *.answers newsgroups in
> > the BI calculations ?
>
> I don't think any FAQs violate a BI standard, but not all systems are
> using BI. Many appear to be using the simpler standard of "less than
> N newsgroups," which does catch lots of FAQs.
Tung-chiang Yang tcyang@netcom.com