Re: [Ietf-languages] Language subtag registration form

[Doug Ewell <[e-mail]>]

I agree with Peter that while the RFCs that define the 'u' and 't' extensions may be used as good examples for future extensions, the code lists themselves should not be. They are wrapped in large zip files along with hundreds of other files, most of which have nothing to do with the extensions, but a few of which do (i.e. you might want the DTDs in order to validate the XMLs where the code lists live).

A better example of how to present the code lists might be ISO 15924:

https://www.unicode.org/iso15924/codelists.html

There are a few HTML tables, nicely formatted for human browsing, but the normative code list is in a single, semicolon-delimited text file. It's zipped to prevent mangling by download protocols ("Oh, good, a text file! Let's change line endings and convert everything to Windows-1252!"), but it's the only file contained within that zip file. Its name includes the release date, and individual records show the date each was added to the list.

At one point I had hoped that the CLDR data items pertaining to the 'u' and 't' extensions would be split out into their own Registry-like files, and I still have examples of how that would work, but in the end, doing this was not a priority of the CLDR project.

To their credit, they are still machine-readable text files, as opposed to HTML that one must scrape (like UN M.49), or any of several native spreadsheet, word-processor, or database formats (as ISO 4217 used to be; thankfully CLDR distills that for us for the 'u' extension). And, of course, they follow all the other requirements of BCP 47: freely available, versioned, stable, etc.

--
Doug Ewell, CC, ALB | Thornton, CO, US | ewellic.org