Re: [Ltru] rechartering to handle 639-6 (was FW: Anomaly inupcomingregistry)

[John Cowan <[e-mail]>]

Randy Presuhn scripsit:

> How much content (as a percentage of internet traffic, or as a percentage
> of on-line library holdings, for example) would be covered by 639-6 (when
> it's done) that would not (or could not) be covered by the registry updates
> recently approved and the normal operation of [e-mail]?

As is hopefully clear by now, ISO 639-6 increases the depth, not the
breadth, of coverage.  Its code elements are a superset of those in 639-3
and 639-5 (represented by the same 3-letter codes), but also include
elements for genetic language collections not in 639-5 and variants
which are more fine-grained than those in 639-3, both represented by
4-letter codes.

Of the latter class, some can be identified with particular BCP 47 tags,
but the bulk cannot due to the lack of corresponding registered variants.
In addition, there are concepts like "Written X" that have no BCP 47
equivalent at present; we can represent "X written in script Y" but not
the concept of "X written in any script."

IMO, adding collection tags based on 639-6 would be a straightforward
extension of the process used to add 639-5 ones; however, doing so
would require activating the currently unused 4-alpha primary language
subtags.  Adding 639-6 variant tags involves identifying those which
cannot be represented by language-script-region triplets and mapping
them syntactically to variant subtags (I have suggested prepending "6"
for this purpose).

-- 
You know, you haven't stopped talking           John Cowan
since I came here. You must have been           http://www.ccil.org/~cowan
vaccinated with a phonograph needle.            [e-mail]
        --Rufus T. Firefly