When I search for Japanese input for Linux, most people refer to Mozc. It works somewhat fine, but it somehow tries to autocomplete words in a weird way, which is not really helpful.
For example when I type みなと, the top suggestion is 港区, followed by みなとみらい and only the third is 港. This gets even worse when I search for more rare kanji at it just gives you more and more suggestions that start with whatever you typed, including katakana words.
I searched the settings, but I did not find a way to disable autocomplete either. Does anyone have another usable IDE for linux? Does this autocomplete thing also happen on windows?
3 comments
The Mozc conversion is not super good by default. You could try to add another dictionary, like UT dictionary to see if that brings it closer to your expectations.
Although your examples are not exactly weird. It’s not super surprising that 港区 and みなとみらい could be more frequently used than 港 by itself. But if it’s weird in general it’s worth a try to switch dictionary.
I use both the Mozc input when I’m on Linux and the windows IME on windows. I’ve found that regardless of the IME you use, it helps to add a bit more context (especially a particle) to the word you’re trying to type to help it figure out what you mean. On windows, if I type みなと I don’t see the 港 kanji until the fifth option. But if I type out みなとに the first option is 港に. Same for 港はどこですか, the only option it gives is the one with the kanji you want. I rarely try to get a single kanji by just inputting the reading anymore, it’s a bit annoying but you get used to it.
Press space instead of tab. That seems to be the solution to most mozc/Google IME-related question that get posted on here
I have [this](https://archlinux.org/packages/community/x86_64/fcitx5-mozc/) vanilla mozc package installed, and when I disable personalization and hit space on みなと the first six options I get are, in this order: 港、みなと、湊、皆と、ミナト、南戸
Tab is for suggestions and usually not that useful if you’re just looking up single words