Avoiding Problems by Avoiding Decisions

Adriaan de Groot adridg at sci.kun.nl
Fri May 14 12:18:07 BST 2004

On Fri, 14 May 2004, Allan Sandfeld Jensen wrote:
> > The only "problem" would be repeated names like
> > but as I said, our policy could be to use the long (complete) name in those
> > cases, since it would be silly to use the short names there.
> >
> > Would everyone agree to choose that policy when naming countries?
> >
> Almost. In case of duplicate names there should be a little "creative"
> flexibility. For instance the two Koreas would make most sense to rename to
> North and South Korea, even if those versions are not on the list.

Every time you use a "creative" naming you're making a judgement call and
making the policy more complicated. So far we have Antonio's original
proposal with two amendments, as follows:

== proposed policy text
We use only the short name (from the ISO-3166 code list) in order to
provide a better non-cluttered user interface, and only use the "long
name" when it's necessary in order to distinguish two countries with the
same "short name". The following exceptions are made:

* The short name of locale KP is "North Korea"
* The short name of locale KR is "South Korea"
* The short name of locales VG and VI is their long name

The short name locales other than KP, KR, VG and VI can be derived from
an entry of the list with the following shell command:

# Assume LONG_LOCALE is set to the complete name as shown in the first
# column of the webpage displaying the ISO-3166 code list:
echo "$LONG_LOCALE" | sed 's+([^)]*)++' | sed 's+,.*++'

This strips bracketed strings (relevant to locales FK, CC, VA as of this
writing) and removes the long name extension after the comma.

== end proposed policy text

Note, by the way, that not all short names are short; Libyan Arab
Jamahiriya isn't short. The names of island locales tend to be lengthy as
well. Locale GS seems to be the longest one.

The advantage of such a policy is (and gosh, it's really close to what
George originally proposed) that it is short, that it is clear, that to
some extent it says "the naming is out of our hands because we accept
such-and-such a list which is considered authoritative". And that's a way
to get threads like this, and discussion in bugzilla, to FOAD.

Note that any decision, even one that defers to another standard, is a
political one and may anger some folks. Using the short name of locale MK
may anger people in another locale who believe that the long name of MK is
misspelled (that , shouldn't be there). I can only hope we can (a) send
the nutcases to argue elsewhere in a body where politicing is the norm (b)
sell this policy as formulated above as 'value free'.

'value free' means that the policy isn't particularly constructed to
produce results beneficial or hurtful to a particular group. For instance,
adding the additional rule:

* The short name of locale WF is "Wallace and Gromit"

would make the policy non-value free (but funnier, for sure - anyone from
locale WF that I meet in future, I owe y'all a beer).

More information about the kde-core-devel mailing list