Analysis of Registrations in the Image Online
Design .WEB Top-Level Domain
Motivation and History | Research
Findings | Research Methodology | More
MOTIVATION AND HISTORY
Discussion following publication of my Analysis of Registrations in the ARNI .BIZ Top-Level Domain suggested that I could perform similar analysis on registrations in the Image Online Design .WEB TLD. I have done so, and these are my findings to date.
As of mid-day Thurday June 21, 2001, there are 22491 distinct SLDs (second-level domains) in the IOD .WEB TLD zone servers. The IOD Whois Server provides no information about dates of registration of IOD .WEB domains. However, it is nonetheless possible to retrieve information from the Whois server about who registered which names, and to request default web pages from each domain. See a table of all IOD .WEB SLD registrant names and page titles (warning: >4MB). Note that 36 domains were both listed as available according to the IOD Whois server and included in the IOD .WEB zone file; these domains are designated "(available according to Whois)" in the table. Note also that a private email from IOD head Chris Ambler asserts that there are 24179 IOD .WEB registrations, and that the IOD .WEB DNS zone file is incomplete: "There are quite a few that have been registered but are not included in the daily zone sweeps for any of a number of reasons. A number of registered domains are not included in the zone because they have specified their own DNS and have had resolution errors."
There is "clumping" of registrations with the most frequent registrants. According to the IOD Whois Server, the top 5 registrants have registered 3228 domains, or 14.7% of the current IOD .WEB SLDs; the top 20 registrants have registered 7342 domains, or 33.5% of the current IOD .WEB SLDs. See table of registration counts by registrant.
A substantial portion of registered IOD .WEB domain names are not currently operational. In my tests, 92 of the 22491 IOD .WEB SLDs had web pages (other than the standard IOD .WEB error message) publicly available at either the http://sld.web or http://www.sld.web address. See table of IOD .WEB SLDs with operational non-IOD error web pages. Another 22357 of the IOD .WEB SLDs display the standard IOD .WEB error message (as shown here) at one or both of these URLs, and also delegate their DNS zone to IOD's default DNS servers according to information in the current IOD .WEB zone file. As those SLDs form the bulk of .WEB SLD registrations listed above, they are not repeated here. All told, 102 distinct IOD .WEB SLDs have delegations in the IOD .WEB DNS zone file to servers other than IOD's default servers. See table of IOD .WEB SLDs with delegations to DNS servers other than IOD defaults.
I used the information at http://www.superroot.com/root.db to obtain the IP addresses of the IOD .WEB TLD DNS servers. (I would have preferred to obtain this information directly from IOD, but their FAQ reflected this information "will be posted shortly." ) I then performed a zone transfer, yielding this file (warning: >800KB) which includes (in addition to certain other content not used in analysis here) a listing of all SLDs currently available in the IOD .WEB TLD.
I used a custom script to remove extraneous content from the IOD .WEB zone file. At the conclusion of this processing, I had a simple listing of SLDs (and corresponding DNS delegations) according to data currently available in the IOD .WEB TLD.
I used a custom script to request the WHOIS information about each domain currently in the IOD .WEB TLD, obtaining WHOIS information from the interface freely available at <http://www.webtld.com>. I rearranged and otherwise analyzed the SLDs, registrant names, and dates of registration to make the various tables and charts reported in Research Findings.
After configuring my testing computer to use the IOD DNS servers as listed in SuperRoot, I used a custom script to attempt to retrieve the default HTML files accessible from .WEB hosts of standard names. When such connections were successful, I stored the title of the resulting HTML page.
Note that all WHOIS data has been left in original capitalizations as provided by the IOD WHOIS system.
This document follows up on similar research done in my separately-posted Analysis of Registrations in the ARNI .BIZ Top-Level Domain document.
I intend this research to spur discussion on related issues. Accordingly, I welcome discussion of the history, research findings, and research methodology, or any other related topics. I may be reached via email at email@example.com.
Other discussion of these research methods and findings is taking place on the DNSO General Assembly mailing list and at ICANNWatch.
This study is my own work and does not reflect any official approval or endorsement by the Berkman Center for Internet & Society. I post it on the Berkman Center's web server thanks to my ongoing affiliation with the Berkman Center, but that does not mean that the Berkman Center endorses my findings in any institutional sense.
Last Modified: June 27, 2001