Comment by muvlon
2 days ago
Meanwhile, Latin a and Cyrillic а, which look identical, have the same origin and are often pronounced identically too, get separate codepoints. CJK unification was not based on any clear preestablished principles but simply an attempt to cling to 16-bit codepoints, Asian languages be damned.
Cling to 16 bit codepoints sounds like a legitimate explanation? Latin and Cyrillic didn't would have saved no relevant amount of space.
It looks like the CJK unified space is over 20,000 characters, so that's a real technical magnitude distinction compared to Latin and Cyrillic. "Asian languages be damned" seems like a bad faith read, compared to "Java and Windows char is 16 bits and that will never change realistically" (and in fact they still haven't, even in 2026 things which rely on UTF16 instead of UCS2 are still commonly bugged unfortunately)
We also have at least 15 copies of Latin alphabet in Unicode, all differently styled/fonted (there is chancery cursive, mathematical cursive, the fraktur version, the full-width version, etc.)
Yes, but most of those are intended to be used exclusively for math, where the different styles do have semantic meaning. Even in introductory textbooks, it's pretty common to see the roman, calligraphic, and double-struck variants of a single letter all used with different meanings.
Plus, these characters were only added 20+ years later, and in the supplemental planes, not the BMP.
2 replies →
That’s correct, it was deemed critical to making Unicode 16bits (although there were other good technical reasons but they would apply to unifying other languages as well.)