3 ms·
Many Japanese websites have migrated from Shift-JIS to UTF-8, but this still ignores the fact that e.g. television captioning uses special characters[1] that ar
by koito17 2y ago
Many Japanese websites have migrated from Shift-JIS to UTF-8, but this still ignores the fact that e.g. television captioning uses special characters[1] that are not found in UTF-8 or Shift-JIS. Windows itself has a habit of using its own Windows-932 encoding, which frequently causes problems in the Unix software I use. (e.g. Emacs fails at auto-detecting this format, and instructing Emacs to use Shift-JIS will result in decoding issues)
[1] section 1.2 in https://www.fontucom.com/pdf/AFSARIBR-Map.pdf https://www.fontucom.com/pdf/AFSARIBR-Map.pdf
- cheeze 2y agoIf Emacs fails to do it correctly, end developers are screwed.
- koito17 2y agoWindows-932 is the text encoding I dread the most. I wish literally everything could use Unicode, but we're not quite there yet. There's a reason encoding-japanese[1] on NPM has nearly 1 million weekly downloads. Unfortunately, I have to use it in one of my React Native applications since the one of servers I speak to returns text encoded as Shift-JIS and Hermes does not implement JavaScript's TextDecoder API. Shift-JIS is marginally better than the Windows code page, but I'd really prefer UTF-8. [1] https://www.npmjs.com/package/encoding-japanese https://www.npmjs.com/package/encoding-japanese
- Dwedit 2y agoAll the television caption characters shown in that chart have a Unicode code-point defined, so they exist in UTF-8.