[ICANN-EU] The Floating Root...ONLINE...

["JIM FLEMING" <jfleming@anet.com>]

If people truely want a bottom-up, democratic system, then root name servers
will not be required. Many ISPs do not rely on root name servers to locate
the
various TLD clusters. If one starts with the TLDs they are interested in
finding,
they can go to some commonly used -DLD.COM names and they can ask those
people where the TLD servers are located for a particular DLD (dash-level
domain). This would be like asking children who their parents are. It works
bottom up.

Here is a small Perl program which will give people a flavor for how this
works.
If you use it with legacy TLDs, it will find the servers in the TLD cluster.
Once
one has those servers more elaborate scripts can be used to cross-check
those
servers. Many ISPs develop their view of the TLD clusters independent of any
root name servers. They load their view in their name servers. This allows
them
to select which TLDs they want to serve. It also makes them immune from
being
dependent on the legacy root name servers (or alternative root name
servers).
Over the past few years, some ISPs have noted that they are MORE stable and
robust and immune from the various root server glitches that have occurred.
By having their own view of the TLD clusters, they stay on top of changes
made
in those clusters.

#!/usr/local/bin/ksh
#
# IPv8 Floating Root DLD/TLD Locator
#
# Usage: locate.sh DLD
#

DLD=${1}

#
# Conditionally make a new directory
#

function Mkdir {
if [ ! -d ${1} ]
then
mkdir ${1}
fi
}

#
# Get the raw nameserver information
#

function getns {
if [ -f ${DLD}/SURVEY/${1} ]
then
return
fi
dig ${1} any | tr "[:lower:]" "[:upper:]" >${DLD}/SURVEY/${1}
grep ' NS ' <${DLD}/SURVEY/${1} >${DLD}/NS/${1}
}

#
# Create directories to store everything
#

Mkdir ${DLD}
Mkdir ${DLD}/SURVEY
Mkdir ${DLD}/NS
Mkdir ${DLD}/TMP
Mkdir ${DLD}/FIND

#
# Query the common places for clues
#

getns ${DLD}-REGISTRY.com
getns WWW-${DLD}.com
getns REGISTER-${DLD}.com
getns DOT-${DLD}.com
getns NIC-${DLD}.com
getns ${DLD}-NIC.com
getns ADDR-${DLD}.com
getns IN-ADDR-${DLD}.com
getns ${DLD}.com

#
# Query the top DLDs
#

# getns ${DLD}-ONLINE.com
# getns ${DLD}-INC.com
# getns ${DLD}-NET.com
# getns ${DLD}-USA.com
# getns ${DLD}-E.com
# getns ${DLD}-UK.com
# getns ${DLD}-WEB.com
# getns ${DLD}-GROUP.com
# getns ${DLD}-IT.com
# getns ${DLD}-DESIGN.com
# getns ${DLD}-SHOP.com
# getns ${DLD}-TECH.com
# getns ${DLD}-WORLD.com
# getns ${DLD}-SOLUTIONS.com
# getns ${DLD}-US.com
# getns ${DLD}-SERVICES.com
# getns ${DLD}-LINE.com
# getns ${DLD}-TV.com
# getns ${DLD}-CONSULTING.com
# getns ${DLD}-1.com
# getns ${DLD}-INTERNATIONAL.com

#
# Build unique list of nameservers
#

cat ${DLD}/NS/* >${DLD}/TMP/ns.all
sed -e 's/^.* NS file://' <${DLD}/TMP/ns.all >${DLD}/TMP/ns.only
sort -f <${DLD}/TMP/ns.only | uniq -i >${DLD}/TMP/ns.uniq

#
# Check the clues for TLD servers
#

echo $(wc -l <${DLD}/TMP/ns.uniq) name servers to check
# cat ${DLD}/TMP/ns.uniq

while read nameserver
do
echo Checking ${nameserver}
if [ -f ${DLD}/FIND/${nameserver} ]
then
continue
fi
dig @${nameserver} ${DLD}. any | tr "[:lower:]" "[:upper:]"
>${DLD}/FIND/${nameserver}
done <${DLD}/TMP/ns.uniq

#
# Collect the potential TLD servers
#

cat ${DLD}/FIND/* | grep "^${DLD}.* NS " >${DLD}/NS.all
sed -e 's/^.* NS file://' <${DLD}/NS.all >${DLD}/NS.only
sort -f <${DLD}/NS.only | uniq -i >${DLD}/NS.uniq

if [ -s ${DLD}/NS.uniq ]
then
echo $(wc -l <${DLD}/NS.uniq) potential TLD servers found for ${DLD}
cat ${DLD}/NS.uniq
else
echo No TLD servers found for ${DLD}
exit
fi
-------------------------------------------------------

Testing the code...

In order to test the above software, you will need some TLDs
that you want to locate. You can obviously use any of the common
legacy TLDs. Other TLDs from the following list can also be
located, IF they exist.

http://www.ntia.doc.gov/ntiahome/domainname/130dftmail/unir.txt

For people interested in looking for the most popular DLDs,
here is a recent survey of the most popular DLDs (with over 5,000
registrations). One can use each name here to locate the TLD
clusters for the following names with the above software.

> 37187 ONLINE
> 15755 INC
> 14787 NET
> 11783 USA
> 9455 E
> 9432 UK
> 8657 WEB
> 8406 GROUP
> 7498 IT
> 7175 DESIGN
> 6935 SHOP
> 6268 TECH
> 6260 WORLD
> 6017 SOLUTIONS
> 5726 US
> 5617 SERVICES
> 5607 LINE
> 5331 TV
> 5212 CONSULTING
> 5188 1
> 5003 INTERNATIONAL

In summary, root name servers are no longer needed. A more stable,
democratic, bottom-up, system can be acheived if one uses the name servers
as "voters" in an election. Here are the results of the above program for
the ONLINE TLD.

16 name servers to check
Checking DNS1.INTERNET-FR.NET.
Checking HOO.FRENCH-CONNEXION.COM.
Checking MY.HOSTING.AC.
Checking NS.CHINA-CHANNEL.COM.
Checking NS.CNET.COM.
Checking NS0.SERVE.CO.UK.
Checking NS0.SERVE.NET.UK.
Checking NS1.BEACHSHORE.NET.
Checking NS1.DOTSTERINC.COM.
Checking NS1.QUASAR.NET.
Checking NS1.ULTSEARCH.COM.
Checking NS2.CHINA-CHANNEL.COM.
Checking NS2.CNET.COM.
Checking NS2.DOTSTERINC.COM.
Checking NS2.ULTSEARCH.COM.
Checking YOUR.HOSTING.AC.
7 potential TLD servers found for ONLINE
CONDOR.TALLSHIP.NET.
FALCON.TALLSHIP.NET.
NOMAD.TALLSHIP.NET.
NS1.HYPA.NET.
NS2.HYPA.NET.
RECTUM.TALLSHIP.NET.
SATAN.TALLSHIP.NET.
--------------------------------------------------

Moving to other topics, such as the Registry and Registrars.
Since the above code uses 9 commonly used names to locate
the TLD servers, it follows that humans might use those same
names from their browsers to try to locate the various people
or companies that serve a TLD Community. All of the 9 names
have been reserved for ONLINE, that is an indication that there
is interest and work progressing.

$ ./check.sh ONLINE
WWW-ONLINE.COM taken
ONLINE-REGISTRY.COM taken
REGISTER-ONLINE.COM taken
DOT-ONLINE.COM taken
ONLINE-NIC.COM taken
NIC-ONLINE.COM taken
ADDR-ONLINE.COM taken
IN-ADDR-ONLINE.COM taken
ONLINE.COM taken

>From here, people can discover that the ONLINE TLD is very
well supported. Via links from the above names, an open and
democractic community can be discovered. People with names
ending in ONLINE, can all participate in making sure that all of
the needed infrastructure is operational. No single-point-of-failure,
top-down, ICANN-like structure is required. As you can see in
this example, we started with the word ONLINE and have
quickly deduced why it has over 37,000 supporters.


Jim Fleming
http://www.unir.com/images/architech.gif
http://www.unir.com/images/address.gif
http://www.unir.com/images/headers.gif
http://msdn.microsoft.com/downloads/sdks/platform/tpipv6/start.asp


----- Original Message -----
From: Gerhard Wendebourg <gwhh@gmx.de>
To: Alexander Svensson <svensson@icannchannel.de>; <icann-europe@fitug.de>
Sent: Sunday, November 12, 2000 8:29 AM
Subject: Re: [ICANN-EU] Interview with ICANN director Helmut Schink


Of course I see, that developing / forcing an alternate root-server-system
is not very easy and simple and will cause some problems and conflicts.
But so we have the chance, to solve these problems under the participation
of the Internet-User-Community instaed of an autocratic institution
established to realize first the interests of the commercial net(ab)use and
the US-Government.

I think, that an At Large Community can effort the ressources to realize
(on the long) a stable working system and its democratic administration,
which the ICANN obviously don't want to serve.
If you look at the developing of Linux you will see, that a non-commercial
/ non-governmental action comes to professional results, that you must not
hide and that can stand  an approval for stable working.

Politically we are now at nearly the same state, the american people were
before there revolution in the 18th century: they were ruled by an overseas
government, that didn't much care about their needs, and so gone into the
risk of chaotic / unstable self-regulating and so born the USA.
There's always a risk of unstability if you get to democratic /
non-hierarchical ruling, but its worth, to go for this - even if you look
at the economical results.


At 13:36 12.11.00 +0100, Alexander Svensson wrote:
>
>Dear Gerhard,
>
>> One of the ways is of course the etablishing and publishing of alternate
>> root-servers:
>> if the At Large Community (?!) decides to support them instead of dealing
>> with an ICANN-Administration, where a user-participation is not really
>> intended, maybe the day will come, when the ICANN looks at an growing
>> paralell (alternate) universe, that is no more object to their
administrating.
>> We would so have an ICANN-commercial-Internet-Community on one side and
on
>> the other an At Large-Net-Citizen-Community from people, that are ruling
>> and administrating themselves.
>> If things like this are really on the way, the ICANN might take a new
look
>> on the situation and think, that Users / At Large-Members are worth, to
>> have some regards to.
>
>in my view, this (alt.roots) would be the very very last and
>desperate resort if ICANN fails altogether. I recognize that
>the notion of being 'alternative' seems to fascinate many,
>but I currently don't see how this really solves any of the
>problems.
>
>First of all, if you have the impression that the
>alternative root servers are free from lawyers, trademark
>problems and the like, I think the main reason for that is that
>few people/companies care about what few people see. Some time
>ago, I calculated that the ASLAN alternative root server gets as
>many querys in a month as the 'legacy' root server A (one of
>thirteen) gets in seven minutes. This relation may have changed
>a bit, but for the majority of Internet users (and trademark
>attorneys), the alternative root servers still are simply
>invisible. As alternative root servers gain popularity,
>companies will try to sue perceived cybersquatters also in the
>alternative name space.
>By the way: E.g. the alternative Name.Space project has a Famous
>Names policy in place and "has already revoked several obviously
>infringing registrations".
>
>Secondly, alternative root servers are only root servers run
>by different people. Whether they are better or worse depends
>on who runs them (and who uses them). Currently, it seems the
>alt.roots are maintained mainly by very small companies, some for
>idealistic, some for more mundane monetary reasons.
>If there was a major decline in use of the 'legacy' (ICANN/US
>Government) root servers, I doubt that it would stay like this.
>To run a worldwide root server system, you need experience and
>money. I would not be surprised to see Microsoft.RootT or
>AOLTimeWarnerDNST -- just pre-configure major browsers, open
>up some interesting sites and start a major campaign. How about
>a EU Root Server System administered by the European Commission?
>All this would be 'alternative', but would it be an alternative?
>
>Thirdly, the tipping over would play havoc with the Internet.
>Currently, there are unique assignments if you stay on the
>'legacy' root. If there were several large name spaces, you
>would have to add the name space to the address -- e.g. I want
>to send a mail to gerhard@wendebourg.nom@icannroot, not to
>gerhard@wendebourg.nom@msroot. Now, if you are lucky, the
>different name spaces are coordinated in some way, so that
>there is only one wendebourg.nom around. But how, according
>to what rules, and by whom shall they be coordinated? What
>happens in cases of multiples claims to a TLD? There would
>be a need for META-ICANN (or rather: the problem would
>have been transferred up one level).
>
>Whichever way we choose, there will always be conflicts.
>I don't believe in any scheme that miraculously makes these
>conflicts go away. Instead, we have to find ways to solve
>these conflicts in a fair manner. Alt.roots are good in
>that they put a bit of pressure on ICANN by showing that
>there are other options, but I doubt that they are the key
>to success.
>
>Best regards,
>/// Alexander
>
>
-------------------------------------
Gerhard Wendebourg          eMail: gw@hwx.de
U&K NetWorX GbR      /       Freelance Networx
WebDesign - Datenbanken - Internet-Projekte
Buergernetz-Hamburg e.V.
fon: +49 40 38613652  fax +4969 791235645
Postadresse:     PF 106104, D-20042 Hamburg