To understand the issues related
to Internet domain names, it helps to start with Internet Protocol (IP)
addresses. This is a number
that identifies each computer (or application) connected to the Internet.
One might think that this is no
different from a telephone number, but this is not quite right. A telephone number (or at least a fixed-line telephone
number) identifies the wire, the line, coming out of a central switch and
going to a home or office. It
does not identify the terminal device (the telephone) itself.
There are other important differences between telephone numbers and
IP addresses, for more details, please see the paper at: http://www.itu.int/itudoc/itu-t/com2/infodocs/015.html
But for the purposes of this very
summary introduction, all we need to know is that each computer on the
Internet is identified by an IP address.
IP addresses are currently 32 bits long.
As the Internet has grown, the number of devices that need IP
addresses has grown, leading to the development of a longer form of IP
address, the IP address Version 6 (IPv6).
A discussion of the issues related to the transition from current
IP addresses to IPv6 is given in the paper referenced above.
But how do we get from IP
addresses to domain names? The
Internet was initially implemented as a research and military network.
As its popularity grew, users and administrators found that it was
inconvenient to have to remember the IP address (a number) of specific
computers or applications, so a system to map user-friendly names to IP
addresses was developed.
There are many ways to build such
systems. The method chosen
was the well-known hierarchical design.
That is, a hierarchical name space was created.
At the top level, some of the names chosen were "edu" for
computers in US universities, "mil" for computers in the US
military, "com" for computers in commercial companies, and
"int" for international organizations (there are several other
so-called generic top-level domain names-gTLDs).
Computers outside the US were under top level codes derived from
the ISO list of country codes, for example "fr" for France
(these codes are called ccTLDs). The
registration of names under ccTLD codes is typically (but not always)
handled by an organization located in the country corresponding to ccTLD
code. This system of mapping
names to IP addresses is called the Domain Name System (DNS).
It is comprised of many individual computer systems.
For more details on the DNS, please see the paper at: http://www.itu.int/itudoc/itu-t/com2/infodocs/001.html
Any number of names can be created
under each name at the top level. The
levels are separated by a period (full stop).
For example, there is an entry "itu" under "int".
Any number of names can be created under second-level names.
For example, www.itu.int identifies
a computer that is the external web server of the ITU.
Whenever a user types the name
(called a domain name) www.itu.int,
software looks up the name in the hierarchical domain name system and
returns the IP address of the external web server of the ITU, allowing
users to access the web pages on that computer.
For more details of how the domain name system is actually
structured and on how it functions, please see the presentation at: http://www.itu.int/itudoc/itu-t/workshop/enum/011.html
There were no real issues with
this hierarchical system so long as the Internet was mostly used for
research. But its growing
popularity and wide implementation led to calls to remove the restrictions
on commercial use that had originally been imposed by the US agencies that
had sponsored the development of the network.
Once these restrictions were removed, there was very rapid growth
in the names registered under the "com" gTLD.
More importantly, the names
started to be used to refer to "real world" objects, rather than
just to refer to computers. For
example, the expectation was that "www.itu.int" would refer to
the web pages of the real-world ITU, and not just to the web pages of some
organization that had chosen to register the domain name "itu",
or to some computer that the ITU had chosen to name "www".
This expectation was particularly strong for commercial companies
registering domain names in the "com" gTLD.
Trademark owners requested that measures be taken to ensure that
their trademarks were not used as domain names by other companies.
In addition, the rapid growth in
the demand for domain names led to calls for the introduction of
competition in domain name registrations, which had previously been
handled mostly by a single organization.
As a result of these pressures, as
well as the recognition that the Internet was rapidly evolving into an
important international infrastructure and that therefore it was
appropriate to revisit the methods that had previously been used to ensure
coordination and stable expansion of the domain name system, a series of
consultations were initiated, which resulted in the creation of the
coordination structure that is in place today.
This and some subsequent developments are described in the
introduction to the proposed draft revision of Resolution 102 (DT 30).