LANGUAGE SUBTAG MODIFICATION - GB

[Doug Ewell]

Lee Gillam <L dot Gillam at surrey dot ac dot uk> wrote:

> Wikipedia - apparently bigger than the Encyclopedia Brittanica - 
> identifies the use of "roa" for Jèrriais. 
> http://en.wikipedia.org/wiki/J%C3%A8rriais I merely reported that 
> those reading such information might use it.

You're correct; that's different from saying that you personally 
advocated it.  I'm certainly one of those readers who sees Wikipedia as 
accurate much more often than not, and probably would have used "roa" to 
tag such content if John Cowan had not steered me away from it.

> The issue I was alluding to previously, and which Peter Constable 
> expanded upon then appears to have retracted but might now want to 
> restate, was related to established user behaviour and "splitting" of 
> identifiers. If, referring to 3166, users see that GB could be used to 
> encompass GG/JE/IM, as per the comment it once had, the combination 
> "fr-GB" seems reasonable. Especially given the "fr" fallback. OTOH, 
> the notion of "British French" may seem ridiculous to some and "roa", 
> while a bin of "other", may be suggested elsewhere and so used. So, as 
> of fairly recently, fr-GG and fr-JE might now be used alongside 
> resources identified by roa and/or fr-GB. And it could take some 
> effort to untangle any previously tagged resources (backward 
> compatibility) based on
> GB(2) = GB(1) - JE(2) - IM(2) - GG(2) [hence GB(2) <= GB(1)]
> and possibly
> roa(2) = roa(1) - fr-GG (2) - fr-JE(2) [hence roa(2) <= roa(1)]
> with some unknown value for roa <=> fr-GB.

These are all good points.  Splitting of region subtags always creates 
sticky problems.  Imagine if, through some stroke of diplomatic genius, 
North and South Korea decided to reunify, and suppose further that usage 
of the Korean language in KP and KR over the past 50 years has diverged 
to the point where the tags "ko-KP" and "ko-KR" indicate a real 
distinction.  If the reunified Korea retained either the KP or KR code 
in ISO 3166-1, that tagging distinction would be lost.

I guess further discussion may be warranted on how best to tag the 
flavors of Norman French used in the Channel Islands.

--
Doug Ewell
Fullerton, California, USA
http://users.adelphia.net/~dewell/