revision of ISO 3166

[Markus Kuhn]

"Peter Constable" wrote on 2004-09-03 21:23 UTC:
> FYI, there is work underway to revise ISO 3166-1. It is currently at the
> committee stage (CD) and is being balloted, with a deadline for comments
> of 2004-9-17. There is a copy of the draft available at
> http://www.niso.org/standards/resources/n419.pdf.

Some quick comments after reviewing the draft (feel free to forward this
to anyone officially responding to the ballot):

Section 5.2:

  "The alpha-2 code uses combinations, in upper case, of two letters of
  the 26-character Roman alphabet (ignoring diacritic signs) from the
  range AA to ZZ."

What ISO/CD 3166-1 calls the "Roman alphabet" is referred to in ISO
10646 as the "Latin script". Looks like the terminology ought to be
sorted out here between the country code and the coded character set
committees.

Section 5.3:

  "This part of ISO 3166 also provides an alphabetic 3-character (alpha-3)
  code ... for use in cases where a specific need has been identified."

It is hard to see, what such a specific need for an extra alpha-3 code
might be, or which specific needs for a 3-letter code this one is
intended to meet. So either the standard should elaborate on what such a
specific need could be, and which ones are taken into consideration when
maintaining alpha-3, or it should drop the alpha-3 code entirely.

[Side rant: I've never seen any application of the ISO 3166-1 alpha-3
code. The International Olympic Committee (IOC) maintains another
alpha-3 country code, which differs substantially from ISO 3166-1 and
uses in some cases the same 3-letter code (e.g., ANT) for another
country [http://www.unc.edu/~rowlett/units/codes/country.htm]. Since
fewer codes for the same application are always better than more, I
suggest to drop the alpha-3 code, and to lobby with the IOC to use the
ISO alpha-2 codes at international sporting events. The latter are now
ubiquitously known worldwide through Internet domain names. Likewise,
the contracting parties to the Conventions on Road Traffic should be
encouraged to consider switching to ISO 3166-1 alpha-2, to reduce the
existing confusion created by having multiple widely-used country
codes.]

Section 7.4.2:

There is a reference to a non-existing section 7.2.2 (should be 7.4.2?)

Section 7.4.5:

Why not include the list of reserved code elements in an appendix?
Why so secretive?

Section 9:

The local short name in the row AZERBAIJAN contains a question mark on
my screen. (Problem with the embedded fonts?)


General issues that should be added:

  - Information about copyright licences for using the code.
    Since there has been a lot of confusion on this issue in the recent
    past, it would be good if ISO 3166 had a very clear clarification that
    these codes and code tables can be used, reformatted and redistributed
    free of charge, without any need to pay licence royalties. [Otherwise
    innocent engineers worldwide will have to waste a lot of time
    talking to corporate lawyers who demand such a clarification for
    due diligence exercises.]

  - It may be useful to know for users of ISO 3166 that it is
    practice to charge the numeric code of a country when the territory
    of that country changes (e.g., Federal Republic of Germany in 1990).
    In practice, the alpha-2 code identifies the shortname of a country,
    while the numeric code identifies the territory of a country
    (which is more useful for collecting statistics). Unfortunately,
    I couldn't find this actually explained anywhere in the draft, and
    I suggest this should be added or clarified, as it is a useful crucial
    detail for some applications.

Markus

-- 
Markus Kuhn, Computer Lab, Univ of Cambridge, GB
http://www.cl.cam.ac.uk/~mgk25/ | __oo_O..O_oo__