Micheal Everson schrieb: >>>> * Add preemptively LATIN [CAPITAL|LOWERCASE] LETTER * WITH CEDILLA >>>> ATTACHED for every Latvian/Livonian character currently in UNicode. >> >>> Why? Latvian and Livonian don't use letters with "proper" cedilla attached. >> >> Maybe my english wasn't perfect here; of course I think that for writing >> Latvian the existing characters shall be used. I meant "for" in the sense of >> "foreach" or "for loop" in programming languages.
>I have no idea what that means. You want to add a bunch of new non-decomposed >characters with a proper cedilla… why? >> And yes, I think not only the four character required for marshallese, but >> also the other ones (g, k, and r). >Why? The first reason is to solve this problem completely and not only to resolve a Latvian-Marshallese conflict and leave some other exceptions for later. The second reason is that the letter g, k, l, m, r with proper cedillas are currently not encodable using UNicode (because of the latvian exceptions and canonical composition/decomposition), but they should *obviously* be encodable. >>>> (Don't use terms like MARSHALLESE [CAPITAL|LOWERCASE] LETTER [M|N] -- such >>>> entities don't exist from a character encoding point of view.) >> >>> Yes they do. Cf. U+0406 CYRILLIC CAPITAL LETTER BYELORUSSIAN-UKRAINIAN I. >>> The character name exists to distinguish it from other characters and to >>> guide the user in the character's use. >> >> But that character exists as a base letter with a distinct shape. There is >> no distinct base letter marshallese m or n. >There is no decomposition. There is no base character + diacritic. The whole >thing is a "letter" used in Marshallese. (It's just a name.) Allthough there is the famous Goethe quote "Namen sind Schall und Rauch" I think good naming style matters, and I prefer the descripte style LATIN CAPITAL LETTER L WITH PROPER CEDILLA (marshallese) to the ad-hoc style LATIN CAPITAL LETTER MARSHALLESE LETTER L WITH CEDILLA. But this is a question of style and can be debatted endlessly without consensus. --Jörg Knappen >Michael Everson * http://www.evertype.com/
