On Mon, Jun 10, 2013 at 10:29 AM, Colosi, John <[email protected]> wrote:
> All, > > > > Per UTR 24, Section > 2.8<http://www.unicode.org/reports/tr24/#Multiple_Script_Values>the COMMON > and INHERITED script values indicate that a code point can be > used with 2 or more other Scripts. But the document is not broadly > explicit about which scripts are compatible with which COMMON/INHERITED > code points. An example in this section indicates that *“U+30FC ( ー ) > KATAKANA-HIRAGANA PROLONGED SOUND MARK is shared between Hiragana and > Katanana [sic]”* and that it cannot be *“used with other scripts, such as > Latin or Greek”*. > It does not say "cannot be used", it says "Neither character is used". UAX #24 and the related properties are descriptive -- they document usage, they don't forbid it. Please also read the following section, "2.9 Script_Extensions Property" In my reading, it feels like the document stops short of saying “U+0660 > must only be used in Arabic and Syriac”. > Unicode does not forbid combinations of characters. > Given that these statements appear as an example, they feel > non-normative. So generally speaking, I’d love to get some guidance about > how registries should treat COMMON/INHERITED code points. Specifically, > should registries impose restrictions on the use of certain COMMON code > points? Is there a document that describes those restrictions, mapping > COMMON/INHERITED code points to a set of scripts? > http://www.unicode.org/reports/tr36/ http://www.unicode.org/reports/tr39/ http://www.unicode.org/reports/tr46/ http://www.unicode.org/reports/tr31/ markus -- Google Internationalization Engineering
