[arin-ppml] SWIPs & IPv6
> I disagreed with Ted yesterday so naturally I have to agree
> with him strongly today. We need whois data that allows us
> to get in touch with a real person in short order.
Now I'm confused. If we need "network contact" data that
quickly gets you in communication with people who are
ready, willing and able to act on your communications, then
why do you call it "whois" data? Whois data is vague,
ill-defined, out of date, and misleading.
We will get further ahead if we drop the discussion of
whois, and improvements to whois, and start from the needs
of network operators and others. I see a strong need from
many parties to know the identity of address block users.
Some people want this just to do long term pattern analyses
of network abuse. I also see a strong need for accurate
and up-to-date contact information that leads directly
to people who are ready, willing and able to act.
However, I don't confuse the two. If a published directory
identifies Joe Bloe of Outer Hoboken as the address block
user, it does not follow that people should contact Joe Bloe.
I'd like to see clear separation of these things so that
people are not surprised to see Chicagoland Hosting Center
NOC as the contact for Joe Bloe in Hoboken. Ask for identity
info and you get one thing. Ask for network admin contact
and you get another.
> In any case when these things happen waiting days for a
> response from an anonymized and infrequently checked email is
> not an acceptable solution. Some times one needs a name and
> a telephone number.
Agreed. The info needs to be up to date which is why I propose
that the recipients of a network allocation should run a
server that allows you to query for this contact info. It is
more likely that an organization will keep its own databases
up to date. ARIN could then either mirror these databases
entirely without human intervention or it could refer queries
to the appropriate server based on the IP address, rather like
an HTTP redirect.
> In some cases having to contact the top level provider and
> then work my way down through a series of book keepers to get
> to a network operator has extended problem times.
It would be quicker if the "working your way down" was handled
by the query tool responding to "redirect" messages from the
> I am not saying IPv6 info will be any better than IPv4 info,
> but I hope it will not be worse.
I'm certain that it will be worse, unless we build something
modern to replace the whois hack from the ARPANET days.