Re: Areas (3): Country lists

In message <01e701c169d8$69703be0$5792fea9@dell400> "Thierry Sourbier" writes:

> Not all 3166 codes represent countries indeed. For example, Saint Pierre and
> Miquelon isn't a Country, but a French territory (It provides France with
> fishes and is used to make colorful stamps :).  As you point out in your
> comment, shall we then consider St Pierre and Miquelon to be part of Europe
> (per Europe->France->PM relationship) or part of North America (per its
> location)? If the latest, then does that mean that France is also part of
> North America?

No, I think that whatever most users think to be most appropriate
would be best for a particular default.

I think that will vary from case to case. The French "BOM-TOM" places
are integrated to a different level in the French economy and
political life than is the integration in some other places, e.g.
those attached to the US, or to Denmark, to Australia, or to
New Zealand, or to the UK.

There would be a case for these to exist in both, but obviously you
can only have one default. The geographic location seems best as a
default to me.

> I think it will be hard to have people agree on a *default*
> Area->country->region relationship even if they know they can override it
> (then what's the point of a standard?).

The obvious point of comparison is ISO/IEC 646-1, which had a
default, which people could register differences from.

That worked very well.


> We will probably need several
> different hierarchies to reflect the different usage (political,
> geographical, economical, trade based, military based, etc...).

I'm sure that's so.

The same
> will probably prevails for area and region code: a single computing code to
> represent all French oversea departements and territories (DOM-TOM) would be
> dearly needed

And could be provided, and/or registered

> but would not make any sens from a geographical perspective
> (as it corresponds to places sprinkled around the globe).

which is why I think it's simplest to have the geographical
groupings as a default.

> To resolve this situation an idea could be to concurrently develop a list of
> code for *referencials* so all the different standards could live under one
> roof.
> 
> The code could be:
> 01 - Geographical
> 02 - Political
> 03 - ...

Fine, as a general model. Should these be incorporated as the 3
letter codes, e.g. as

A22 Northern Europe but
A2A European Union
A2C Council of Europe
A2H Organization for Security and Cooperation in Europe
etc,

or as additional subtags, e.g.

A21 Europe
A21-EU - European Union, etc?

> and why not:
> 15289 - World as Company X views it (so we can create mapping tables for all
> the legacy locales).

Indeed, why not?

> One could register areas, regions and relationship based on those
> referential. The referential list would of course need to be pretty open yet
> standardized so mapping table could be written (e.g. the political code for
> the French territories and departement region should be mapped to all the
> geographical regions codes).
> 
> That seems like a huge task but of course may be I missed a simpler way to
> handle the complex nature of such classifications and/or may be my mind has
> drifted from the purpose of such a classification...
> 
> Regards,
> Thierry.

Best regards

John Clews

--
John Clews,
Keytempo Limited (Information Management),
8 Avenue Rd, Harrogate, HG2 7PG
Email: Webbing@sesame.demon.co.uk
tel: +44 1423 888 432;

Committee Member of ISO/IEC/JTC1/SC22/WG20: Internationalization;
Committee Member of ISO/TC37: Terminology

Received on Sunday, 11 November 2001 16:42:55 UTC