w3c / i18n-activity

Home pages, charters, style-guides, and similar documents related to the W3C Internationalization Activity.
65 stars 22 forks source link

Country code reference is wrong. #192

Open aphillips opened 8 years ago

aphillips commented 8 years ago

https://www.w3.org/Mail/flatten/index?subject=https%3A%2F%2Flists.w3.org%2FArchives%2FPublic%2Fpublic-dwbp-comments%2F2016Jul%2F0032.html&list=public-dwbp-comments

original version of comment:

https://www.w3.org/TR/dwbp/#ReuseVocabularies

Example 15 contains this citation:

The Library of Congress publishes lists of ISO 639 country codes as Linked Data (see [ISO639-1-LOC] for two-letter codes):

ISO 639-1 is a list of language codes, not country codes. The standard for country codes is ISO3166-1.

If language codes are desired as an example, ISO 639-3 is a better source for language subtag information (for various reasons having to do with how ISO639 is currently maintained). However, citing ISO3166 instead of ISO639-1 would be preferable here.

fsasaki commented 8 years ago

Unfortunately the agency hosting ISO 639-3 does not yet provide linked data = URIs for language codes. I agree that the text should say language codes. But I don't see an alternative for the LOC based URI in

:stops dct:language http://id.loc.gov/vocabulary/iso639-1/en .

aphillips commented 8 years ago

@fsasaki But ISO 639-1 only covers the alpha2 codes. It's not that helpful as a result. Is there a suitable CLDR reference instead, perhaps? I'm not aware of one for BCP47's registry, which would actually be ideal here.

fsasaki commented 8 years ago

But ISO 639-1 only covers the alpha2 codes

I know. That is a missing bit in the language code community = standardised identifiers for all other ISO 639 part.

Is there a suitable CLDR reference instead, perhaps

I don't know.

I'm not aware of one for BCP47's registry, which would actually be ideal here.

Yes, yes, yes :) If you see a way to convince the IETF to host a linked data based registry, everybody would love it. Or if IETF can't do it, somebody else who is accepted by the BCP 47 community. That would solve a lot of problems in the dwbp review.

aphillips commented 8 years ago

Oh the violent agreement! :-D

Well, IANA could maintain one for BCP47. We (for some value of "we") would have to write an RFC to have IANA convert the cookie-jar format the registry uses to a LinkedData format and maintain appropriate URLs for the registry. I shudder to think of that as being a revision of BCP 47 itself.......

fsasaki commented 8 years ago

I shudder to think of that as being a revision of BCP 47 itself.......

Yes, such a huge process should be a avoided.