From: "Marion Gunn" <[EMAIL PROTECTED]>
> I ask the patience of the Unicode and IETF-L moderators for now posting
> on their lists this request for contact details for the ISO 3166 mailing
> lists (if any). 
> 
> Context: Ireland advisability of reserving 'EI' tag for cited usage
> (baggage-handling at international airports)  and the fact I am only
> discovering now that some things such as this, which I had taken for
> granted as being registered/reserved for/by NSAI have yet to be so
> registered/reserved.    
> 
> For this I blame only my own failure to discover this until lately. 
> Thanks,

If you look within the ISO3166-2 (which is the registry of subcountry codes, for 
regions, states, departements, emirates, counties, etc...) you'll see that Northern 
Ireland is not strictly listed. Instead the registry contains codes for the counties 
located in Northern Ireland.
Unlike most (all?) other 3166-2 codes, their length is usually long (up to 8 
characters) and contain a space after the leading prefix "CO" used only for counties 
in Northern Ireland, but not for counties located in England, Wales or Scotland 
(commonly refered as Great Britain or just Britain).
Shamely, the official ISO3166-1 code for United Kingdom is GB, not UK which is just a 
IANA assignment, both of which include Northern Ireland, and also other UK 
dependancies (but only in ISO3166-1, because IANA defines separate codes for 
dependancies and overseas areas having a local form of governance with semi-autonomous 
status within the United Kingdom, exactly the same way it occurs for dependancies of 
France, USA, China, Chile, Peru).

Using ISO3166-2 codes requires encoding also the ISO3166-1 part (i.e. the 2-letters 
country code). Shamely, ISO3166-1 also contains legacy/historic codes for dependancies 
(for example: The French department of Guadeloupe, or the US state of Puerto Rico).

My be, the simplest way to encode Northern Ireland according to ISO3166 would be to 
use the country code (GB), followed by a code matching all counties in Northern 
Ireland (CO). This would give the quite uncommon code GB-CO!

ISO3166-1 has its known problems (even if it's still better than FIPS which forgets to 
encode many areas, or that uses codes specific to a US government usage, and does not 
match any code used in the referenced country), ISO3166-2 is even worse (many errors 
and omissions, but still much less than FIPS which is very incoherent!)

There's no clear solution for you, so if you need to use a code in a delimited context 
(such as baggage registration in airports), the best way is to use a code that matches 
the uses in the air sector (for example the international codes for airports, which is 
an abreviation of the city name or the name of one of the airports for that city, such 
as JFK for New York).

The airport code is certainly the most appropriate for your usage, and it can be 
correlated uniquely to a single ISO3166-1 country code (forget the definitions of ISO 
3166-2, which is too much incomplete, unstable, unmaintained, still lacks a policy as 
i is maintained only by relevant countries that forget to update their records, and 
contain many typos or simply errors, and outdated as it refers areas that simply don't 
exist on a legal base!).

You won't find these codes in ISO3166. You need to find reference from the 
international air regulation authorities.

Alternatively, you could use the United Nation numeric codes used for statistics 
reports, and that are very well maintained (needed because accurate statistics are the 
common base for international negociations and diplomacy, and these take into account 
common area divisions or groupings such as Northern Ireland, or European Union, or 
OECD countries, or members or parties to international alliances or treaties such as 
NATO).

If you need samples of these UN 3-digit codes, look at the many statistics and reports 
published on the UN web site. Some reports require paying a small fee to contribute to 
UN activities. The UN reference list can be ordered (look on the website for details).

-- Philippe.

Reply via email to