Ubuntu Linux (18.04LTS) using Mozc input method. Typing in Japanese in a LogSeq block: no text is echoed as one types a word. In Ubuntu’s simple text edit application (Gedit), Japanese text input works as a Japanese typist would expect: If I’m typing the word gakusei (学生), I type ‘g’ and the ‘g’ is echoed on the text line where I’m typing. ‘a’, and “ga” corresponds to the hiragana character “か”, so that’s what appears (along with a pull-down menu of the system’s best guesses at the kanji character or characters I’m trying to type). ‘k’ and it echoes the ‘k’. ‘u’ and it echoes the hirigana character ‘く’ since “ku” corresponds to this character.
In LogSeq, nothing is echoed on the text line on which I’m typing, so I only have the top line of the pulldown menu (the system’s context-dependent best guess at what character(s) I’m trying to type) for feedback as to what I’m typing, which is a bit confusing, and not what a Japanese typist would expect (using the Japanese input method of MS Windows – sorry – for instance).
If I type in Japanese/mozc mode in the web app Google Docs, it behaves the same way as Gedit (echoing characters as expected). Interestingly, if I type in LibreOffice Writer, it fails to echo in the same way LogSeq does, so I can’t rule out the possibility that the behavior results from interactions between the application and OS APIs that would be difficult to code around. It’s also a bit surprising that I can’t find any other forum post mentioning this; it’s definitely an annoyance but the app is still basically usable. I’d be interested in what Japanese users of LogSeq think about this.