language subtag registration request: Region EU
mark.davis at icu-project.org
Thu Jan 19 18:34:00 CET 2006
This is not really a tenable goal. There are a couple of hundred false
negatives (as you call them) for essentially every language already in
3066. Does ja-AQ differ substantially from ja or ja-RO or ja-RU? The
addition of the region codes, and the addition of EU, does not alter
that in any material sense; they are really drops in the bucket.
Instead, the goal is to provide people with the tools to express
distinctions that are important in their domain of interest, but to
caution people not to make unnecessary distinctions in a broader domain.
And in a narrow domain, there very well may be some use to
distinguishing ja-RO from ja-RU.
Jeremy Carroll wrote:
> Peter Constable wrote:
>> Is there a real scenario in which a distinction is made -- something
>> like de-EU vs. de-DE, or fr-EU vs. fr-FR? I don't
> > believe Markus has demonstrated such distinctions exist.
> In the Semantic Web applications that I deal with there is cost in
> false negatives, i.e. de-EU appearing different from de-DE while in
> fact being the same.
> If in fact xy-EU was always equivalent to xy-XY where xy is a language
> spoken in the EU and XY is the country most prominently linked with
> xy, then the lack of syntactic equality between the two tags would
> cause problems that would need code. I would like to be clear that
> there is benefit in return for the cost of that code.
> (Admittedly 3066bis, by allowing the use of the region codes, does
> expand the problem of false negatives significantly over 3066, and I
> am not expecting to have to deal with it in practice; but that
> extension was at least motivated by one or more compelling examples of
> Ietf-languages mailing list
> Ietf-languages at alvestrand.no
More information about the Ietf-languages