On Tue, Mar 30, 2004 at 04:03:23PM -0800, Hallam-Baker, Phillip wrote:
I view this as bad. We must provide a means by which people can ease
into the use of MARID. One such mechanism is the ability to define a
grey area between "absolutely accept" and
"probably/absolutely reject".
You mean 'dump it into the spam filter with a large penalty' ?
Actually, I think that would be appropriate for the "probably reject"
category. Equivalent to ~all. But there needs to be something
equivalent to ?all as well; something that says, "the only thing known
about this message is that there's nothing known about this message."
I think there are three cases of interest
1) There is no profile information for the domain
2) There is profile information for the domain and the use is consistent
3) There is profile information for the domain and the use is not consistent
I agree with #1 and #2, but I don't believe #3 is of interest, because
of the use in which records are dynamically updated. There should be no
penalty (or even extra scrutiny) just because a record is less than $x
time units old, unless we're proposing some mechanism to assess and
track behavior (as measured by message content), as well.
--
Mark C. Langston Sr. Unix SysAdmin
mark(_at_)bitshift(_dot_)org
mark(_at_)seti(_dot_)org
Systems & Network Admin SETI Institute
http://bitshift.org http://www.seti.org