What is unk_v? about whosonfirst-names HOT 6 CLOSED

from whosonfirst-names.

A few comments:

• unk_ looks like a Geoplanet-ism and should really be und_ or undefined, per this:

http://www.loc.gov/standards/iso639-2/php/English_list.php

That und_ could have a suffix (preferred, variant, etc) is a bit weird on the face of it since you know... undefined. One thing at a time.

• Where appropriate, things in unk_ or und_ should be placed in their correct language buckets.
• Currently there are no deviations from the basic Geoplanet convention but that doesn't mean there aren't errors (or lapses) of judgement in data imported from Geoplanet (Suomen Tasavalta vs. Suomi)
• Currently we only define ISO-639-2 language codes, but we will shortly be incorporating support for script subtags:

http://www.w3.org/International/questions/qa-choosing-language-tags
http://www.w3.org/International/articles/language-tags/
https://tools.ietf.org/html/rfc5646

The convention, historically, has been to use two-letter codes for country/subtags notation but given the fact that we are already using the Geoplanet syntax we may just stick with three letter codes. It will all be documented though :-)

Finally, this is a good place for syntax/convention specific issues or questions.

from whosonfirst-names.

Related:

https://github.com/whosonfirst/py-mapzen-whosonfirst-languages#mapzenwhosonfirstlanguagessubtags

I suspect what we will do is follow the lead of the Unicode Consortium and publish a "conformance" and "conversion" document:

http://www.unicode.org/reports/tr35/#BCP_47_Conformance

The only question (I think) at this point is whether to expand the "_p/s/v" suffixes in to use "x" (subtag) extensions with a fully qualified label (preferred, colloquial, variant) ...

from whosonfirst-names.

Also this (which is not being implemented anywhere yet):

https://github.com/whosonfirst/py-mapzen-whosonfirst-names

from whosonfirst-names.

I'm going to close this issue, as this should no longer appear in the Finland record, or any other record in Who's On First.

Name properties should now follow one of the following conventions:

name:{iso lang code}_x_preferred
name:{iso lang code}_x_variant
name:{iso lang code}_x_historical
name:{iso lang code}_x_colloquial
name:{iso lang code}_x_unknown

See also: https://github.com/whosonfirst/whosonfirst-properties/blob/master/properties/name.md

from whosonfirst-names.

Hi, I am dealing with UAE localities at the moment, and the following name key comes up very often:

name:unk_x_variant 

e.g. Abu Dhabi or ضدنة

According to the IANA' language subtag registry, the unk language subtag refers to the language spoken by the Enawené-Nawé people.

Provided the frequency ot the name:unk_x_variant key and the fact that the Enawené-Nawé language is spoken only by a handful of indigenous people in the Brazilian rain-forest, it seems unlikely that the unk language subtag refers to the above language.

Am I safe to assume that unk is the unknown value inherieted from Geoplanet?

from whosonfirst-names.