ISO 3166

Infoman Inc. mpereira at
Wed Feb 23 23:49:31 UTC 2000

When it was agreed, almost a decade ago, that ISO 3166 should be split into
a three part standard, the intention was that Part 1 would contain only
"countries", i.e.nation-states, Part 2 Subdivisions was supposed to contain
all entities which are parts of countries (e.g.dependencies, etc.) as well
as anomalies such as territories, trusteeships, etc. which various Part 1
countries administer on behalf of the UN and/or which have a
special/peculiar relationship with a country (= full member in good standing
of the of the UN , e.g. Taiwan  no longer is). ISO 3166-3 focuses on

Unfortunately this "rationalization" did not happen. Too much "politics".
One result is that ISO 3166-1 continues to contain codes for entities which
are not countries. As a matter of fact some the "same" entities appears two
or even three times with different codes (check out "St.Pierre and Michelon"
French islands of the coast of Newfoundland.

When the IP addressing schema was established, ICANN simply referenced ISO
3166-1 country codes believing that it contained only countries (top level
domain names did not have the importance significance they later turned out
to have.) We continue to suffer.

A new ISO/IEC standardization work item has been approved to address the
problem of existing code lists containing entities with different attributes
(or instances of several different object classes).It is titled
"Identification, Mapping and IT-enablement of Existing Standards for Widely
Used Encodable Value Domains".

This NWI is being started via ISO/IEC JTC1/SC32 Data Management and
Interchange, WG2 -Working Group 2 Metatadata. I will be its project editor.

If you interested in this work and similar problems with existing code sets
used in commerce/e-commerce which are "organically grown". Contact the
ISO/IEC JTC1/SC32/WG2 group in your country (e.g. via your national
standards association) [or as a last resort contact me]  - Jake Knoppers
<<mpereira at>

Original Message-----
> From: Law, Gwil Jr. [mailto:Gwil_Law at]
> Sent: February 22, 2000 10:51 AM
> To: Timezone mailing list
> Subject: ISO 3166
> Joseph S. Myers wrote:
> > ... Now I might wonder whether the ISO 3166 criteria for
> > dependent territories are documented somewhere?
> There are some documents.  The best place to find them is at the home page
> of the ISO 3166 maintenance agency,
> <>.  The
> link called
> "How to get a country name into ISO 3166-1" is informative.
> The standard itself says, "The list appearing in this
> International Standard
> ... is based on the list included in the 'United Nations Standard
> Country or
> Area Code for Statistical Use' established by the Statistical
> Office of the
> United Nations....  Some additional entities are included in this
> International Standard with a view to providing more complete global
> coverage."
> The "additional entities" at this time are Antarctica and nine islands or
> island groups, mostly in the Indian Ocean.  Conversely, Channel
> Islands and
> Isle of Man are on the Statistical Office list but not ISO 3166.  You can
> find the Statistical Office list at
> <>, along with a
> disclaimer
> as to the legal standing of the countries.
> A few years ago, a new code was assigned:  FX for Metropolitan
> France.  The
> standard cautiously refrains from defining the scope of each code, with a
> few exceptions, but the apparent intention was that FX would be just
> European France, including Corsica, whereas FR would
> theoretically cover the
> departements and territoires d'outre-mer.  The FX code was later
> withdrawn.
> Despite appearances, this doesn't change the meaning of FR:  the standard
> has always allowed you to use either MQ or FR for Martinique (a
> departement
> d'outre-mer), according to your information interchange requirements.
> When Hong Kong became a special administrative region of China in
> 1997, its
> code and name were allowed to remain unchanged in ISO 3166.  The
> ISO 3166/MA
> Annual Report for 1998 gives three reasons:  "1.  China [has] neither
> requested Hong Kong's removal from the standard nor a change of
> its present
> name form.  2.  ...the [existing] code elements ... are very widely used.
> 3.  According to the UN Statistics Division the new name is to be used in
> statistical publications.  ISO 3166-1 is not a statistical publication."
> I've heard that the ISO 3166/MA receives numerous requests for new code
> assignments that may have no geographic meaning, but are wanted as new
> top-level domains.  The agency turns them down.
> Antoine Leca wrote:
> > BTW, does someone know if ISO 3166 text is or will be available
> > online or for a modest/reasonnable fee?
> I've always heard that ISO is very protective of its copyrights, claiming
> that the only way it can support its activities is to charge high
> prices for
> its exclusive documents.
> Yours,    Gwillim Law

More information about the tz mailing list