Preferred Values for Irregular Tags

[Phillips, Addison]

> 
> > %%
> > Type: grandfathered
> > Tag: cel-gaulish
> > Description: Gaulish
> > Added: 2001-05-25
> > ***Preferred-Value: xtg // most likely Gaulish being meant
> 
> Which Gaulish is xtg?

It's Transalpine Gaulish.

> 
> > %%
> > Type: grandfathered
> > Tag: en-GB-oed
> > Description: English, Oxford English Dictionary spelling
> > Added: 2003-07-09
> > ***Preferred-Value: en-oxedict // with proposed new variant
> 
> oxford would probably be better. But why en-oxford, and not en-GB-
> oxford? Does it matter in this case? (There could indeed be en-US-
> oxford.)

Since the 'oxford' subtag isn't actually limited to 'GB', it might not be useful to included it in the Preferred-Value. Another way to say this would be: what would the Prefix field be in the record for variant subtag 'oxford', 'en' or 'en-GB'?

> 
> > %%
> > Type: grandfathered
> > Tag: i-enochian
> > Description: Enochian
> > Added: 2002-07-03
> > ***Preferred-Value: xxx // Ask for new code from 639-3. This is a
> > bizarre invented language, but there is precedent for invented
> > languages.
> 
> Erm, how invented is this? It's hard to find an actual grammar or
> wordlist longer than 700 lines.

It may not be fully invented. Alas, it is registered here. Another artificial, slightly-weird solution would be variant 'enochian' with prefix 'art' (Artificial languages). But it seems better to me to ask ISO 639-3 first. If they declined, we would also be in a position to register a longer-than-4-character primary language subtag ('enochian') to replace the i-flavored tag.

Addison

Addison Phillips
Globalization Architect -- Lab126

Internationalization is not a feature.
It is an architecture.