CircleID

great article, tom - i've just linked to it from that mega thread that geoff's article started off.

Facts to dispel pernicious addressing myths - this article is an excellent addition to CircleID. If only this could be mass-emailed to IPv6 proponents and ITU participants :)

Great job, Tom.

Please see the latest 200 report for the IP address. Top 11 countries dominated 80% of the allocated IP address ranges.

http://www.ip2location.com/ip2location-internet-ip-address-2008-report.aspx

Tim Ruth said:

Please see the latest 200 report for the IP address. Top 11 countries dominated 80% of the allocated IP address ranges.

http://www.ip2location.com/ip2location-internet-ip-address-2008-report.aspx

Hi Tim,

Thanks for the link. However, people that follow it should remember a couple of things:

The cumulative allocation record you're referencing is just that: cumulative. It depicts the aggregate results of all IP address transactions since the beginning of the Internet. So, in this sense, the view you recommend is a bit like looking at the absolute number of all television sets sold to-date at a point in time thirty years after TV was invented. In some countries, TV has yet to arrive; elsewhere it's recently become available, but is affordable only by economic elites. And in some countries, some well-off households may be on their second, third, or fourth, or later generation of TV (with the older TVs languishing in storage, or maybe the dump), and/or may currently have more than one in their home/place of work, etc.

Compounding the general "historical bias" to this approach is the specific development trajectory of IP and Internet routing technologies and practices, which were vastly less efficient in terms of IP address utilization back when the Internet was conceived as just an experimental platform linking a few universities (almost all of which were in the US). So, in terms of the single-year transaction record suggested above, the allocation records for the years 1983-1990 would show very few transactions per year, but each involving a very large number of IP addresses. In each subsequent "era" of technology and address allocation rules (e.g., 1991-1994, 1995-1997,1998-present), that ratio of IP addresses per transaction would fall—at the same time that the number and geographic diversity of allocations would rise dramatically.

Bottom line: If you were to look at the annual allocations for more recent years, you would still see a substantial bias toward wealthy countries (where there are more prospective paying subscribers and more "local" content), and countries with diversified, competitive telecom markets (where there are more IP address allocation seekers—including "foreign" content providers that have fled markets with artificially inflated prices—and where lower subscriber prices permit users to spend more time online, thus requiring a higher user/address ratio, etc.), but the distribution would be vastly less skewed than the cumulative historical charts presented on your referenced site.