ISO 639-3 changes: batch 2

[Doug Ewell]

> I could look up how often we have done this.

I counted roughly 16 instances since 2010 (not including "Sanapaná"
this year) where ISO 639-3/RA has split one language into two or more,
one with the original name, and where we have deprecated the old subtag
and added new ones without any special action to disambiguate any
Description fields.

In four of these cases in 2012, I didn't even use the word "proper",
which should have made things even more confusing:

"This registration tracks a change made to ISO 639-3 effective
2012-02-06, splitting the code element 'wiw' for Wirangu into two new
code elements, 'nwo' (Nauo) and 'wgu' (Wirangu)."

I believe the following passage in RFC 5646, Section 3.1.5, was added
partly to help explain situations like this:

"To help ensure that users do not become confused about which subtag to
use, 'Description' fields assigned to a record of any specific type
('language', 'extlang', 'script', and so on) MUST be unique within that
given record type with the following exception: if a particular
'Description' field occurs in multiple records of a given type, then at
most one of the records can omit the 'Deprecated' field."

This means the Registry can have two subtags with the same Description,
such as "Sanapaná", as long as one of the two is deprecated. In the
present case, like the 16 before it, the old "Sanapaná" is deprecated,
and BCP 47 recommends against generating it, but allows it to retain its
old meaning of "Sanapaná" for existing content -- even if "Sanapaná"
at the time also encompassed what is now thought of as a new language
called "Angaité".

--
Doug Ewell | Thornton, CO, USA
http://ewellic.org | @DougEwell