addison at amazon.com
Wed Sep 24 22:15:10 CEST 2008
> es-US is going to be very distinct from es-SP
Definitely, since 'SP' isn't a valid region code :-).
> As for the ongoing discussion, there is data tagged zh-TW, and when
> that data is romanized into Hanyu Pinyin, it will be very natural
> tag it zh-TW-pinyin, and I see it hard to argue that's formally
> So zh-TW(-Latn)-pinyin can't be trusted to be Tongyong Pinyin.
No. I think that's the point Randy is trying to make. A subtag that means, ambiguously, just 'pinyin' can have an implied additional meaning derived from surrounding subtags. But that derivation may be wrong. An explicitly defined subtag isn't ambiguous (within the limits of what it defines).
So "zh-Latn-TW-hpinyin" is definitely not Tongyong (ditto a "zh-Latn-TW-pinyin-hanyu"). Whereas "zh-Latn-TW-pinyin" might be either--or some other pinyin.
What is key here for me is whether this distinction is actually important for the requesters or others who would use the subtag. If, in practice, the only thing that matters is "it's in Latin script and it isn't Wade-Giles", then 'pinyin' does that just fine.
More information about the Ietf-languages