2011/8/31 Doug Ewell <[email protected]>: > > Or you could actually follow BCP 47, and use "x-www" instead.
No, because locale tags in BCP 47 starting by the "x" singleton subtags are not parsable to differentiate a language, a region, and a script (as well as other Unicode "u" extensions). They just identify a locale as a whole. The advantage of using a singleton subtag like "q" would be that it would be followed by any subtag giving its private-use value as a language, and it would also not even require any further reservation or private use areas for the future 4-letter codes. It would immediately offer a very cumfortable space (possibly using 1 to 8 letters for the second subtag following it) for any very simple algorithmic mapping of subtag pairs used privately for identifying languages, but all the rest of the locale tag would remain parsable and would keep its structure (the cumfortable space would easily avoid collisions). This is of course an architectural change, but with minimum impact. Anyway, it's highly probable that if ISO 649 starts allocating 4-letter codes, it will also include the space qaaa-qtzaa as a private use area (just like the existing spaces qaa-qtz and qa-qt). But let's not anticipate what will be in this may-be future extension of ISO 639. The "x" singleton space of BCP 47 will then remain for applications that don't want to identify specific languages; it is clearly not ideal to use this "x" singleton for identifying languages but not for the other separable identification properties contained in a locale identifier. -- Philippe.
