3 ms·
The codepoint mentioned in the article, U+5203, is a perfect example: in most of the world it will display in Simplified Chinese, whereas there's a counterpart
by lmm 3y ago
The codepoint mentioned in the article, U+5203, is a perfect example: in most of the world it will display in Simplified Chinese, whereas there's a counterpart U+2F81E that (in practice if not in theory) can be used to reliably display in Traditional Chinese. Only the Japanese rendering doesn't have a codepoint and is forced to use out-of-band methods of indicating the desired appearance.
- adastra22 3y agoIMEs do not and should not select U+2F81E, however. That core point exists merely for obscure Unicode compatibility requirements.
- lmm 3y agoThat's the Unicode consortium's opinion. Many people believe they're wrong about semantics and that Han unification was a mistake. Regardless of the Unicode consortium's intentions, these compatibility codepoints allow one to write text that displays Traditional Chinese correctly without additional out-of-band information - but there is sadly no equivalent for Japanese.
- adastra22 3y agoExcept then you get those characters rendered as little squares for Unknown Char on a wide variety of platforms. There’s good reason to stick to the basic multilingual plane.
- lmm 3y ago> you get those characters rendered as little squares for Unknown Char on a wide variety of platforms. That can happen in the BMP too. It's a well-known problem with well-known solutions. (E.g. you can ship fonts that support all the codepoints you use).