At 12:34 AM 8/6/2003, Margie wrote:
--On Tuesday, August 05, 2003 5:05 PM -0700 "Hallam-Baker, Phillip"
<pbaker(_at_)verisign(_dot_)com> wrote:
>
> This action is a regular occurrence, not a possibility.
There are over 400 DNSBLs that *I* know about. It would be completely
impossible for it to happen, even ONCE on the vast majority of them.
The criteria for listing on most lists are yes/no bits, such as "is
it an open relay?", "is it an open proxy?", "is it a dynamically
assigned IP address?", "is the IP space in the USA?", "is the IP
address in China?"... "does this mailing list do confirmations?".
Margie, I believe that Phillip's argument applies solely to RBLs based on
people voting - "community based" RBLs, which are been discussed. However,
RBLs like all other tools, are susceptible to errors and mistakes, and it
is possible that occasionally an erroneous listing will creep in. It is
also possible that an insider or someone who is trusted by the operator of
the RBL, can list an innocent site, whether by mistake or on purpose. There
is also an issue of different policies used by different RBL operators.
HOWEVER, a consent based anti-spam system that utilizes information from
many sources, and does not rely solely on a specific RBL, can cancel out
the impact made by this mistake, by using other RBLs and sources of
information. It also leaves the choice to the developers and
administrators, to pick and choose which RBLs to use based on their own
needs and wants. This way we do not have to get involved into deciding what
kind of proper philosophy should an RBL have.
On a related note, would you, Margie, and perhaps other RBL operators, want
to help in writing BCPs for RBL operators, and mail administrators that use
them? I am sure there are many technical issues which can be summarized in
a BCP (and many political and philosophical RBL issues which can be left out).
Yakov
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg