[ppml] Last Call for Comment: Policy Proposal 2003-3

[Michael.Dillon at radianz.com]

>"Ah, that'd do nicely, sir - we can even reduce the funding for our AUP 
team
>without anyone being the wiser"

You seem to be saying that an ISP could be colluding with
spammers by ignoring any abuse complaints about the spammer's
IP addresses. This may indeed happen but it is not ARIN's
job to do anything about it.

ARIN is still publishing the contact information of the 
colluding ISP. They may be able to ignore complaints from
3rd parties but they won't be able to ignore complaints
from their peers or their customers. It is trivial to
identify the peers of a rogue ISP in order to file a complaint
about the rogue. 

This all comes back to the purpose of the whois directory.
It all started back in the days of the ARPAnet when someone
decided that anyone who had a user account on an ARPAnet
system should be listed in a central directory for some
reason, probably related to funding for the network connections
and the timesharing systems. The world has changed a lot
since then and nowadays we seem to have no agreed purpose
for the whois directory other than tradition.

In my opinion, the directory is to identify contact points 
for people who can deal with networking issues like
network abuse and misconfigured devices that cause problem
for other people. And if that is the purpose, the directory
should *ONLY* contain information for contacts that are
ready, willing and able to receive communications about 
networking issues and act upon them. If an organization 
cannot meet that basic requirement, whatever the reason, 
then their information should *NOT* be listed in the 
directory.

All the arguments that I have seen about whois are making 
unstated assumptions about the purpose of the whois
directory. I think that is bad because we end up arguing
at cross purposes about different things. We need to begin
by agreeing on the foundation, i.e. what is the purpose
of whois?

I might be persuaded that whois should also serve some
research or network forensic goal. Perhaps we really
should publish all assignments down to a /29 level with
a class-of-user and a city-of-service and city-for-billing.
Some people would find it useful to distinguish between
addresses used by residential users, companies, non-profit
organizations, schools, etc. Some people would find it
useful to know that the city-for-billing of address blocks
used for spamming mostly clusters in a few counties of 
south Florida. But in order to do this we need to
make a clear distinction of when contact information
should be published and when only research information
should be published.

--Michael Dillon