I’m starting to find using a Japanese ime relatively bothersome.
Mine (not sure about other ones) has a feature that shows recently used ““words”” higher on the list. This is certainly useful, but it has one large caveat. It tends to show me whole phrases that I’ve used in the past. This pushes all the actual words I want to look for down. Wanna find 人, well, good luck, it’s now on the second page.
Maybe it has to do with how I break up sentences into conversions. I tried starting with individual words, that still included particles after the conversion suggestions, so I decided to play along and do word + following particle. This sometimes has amusing side effects, but it works for the most part until you get particleless words.
So how exactly do you use your ime and do you want to throw your computer out the window each time you do?
How are you writing 人 that it doesn’t show up on top?
The only issue I have with Japanese keyboards is it not turning some words into kanji or viceversa when writing a full sentence.
I write 人 and then I get a bunch of phrases listed, kinda lame
I type ひと, press space, it presents me with 人, and then I hit enter.
What IME are you using? I’m using the default Microsoft IME.
It’s a Linux one, it’s called fcitx5, but tbh, it’s not a problem I won’t be able to solve, I’m just curious if it’s a problem I need to solve with the program itself, or if it’s only because I’m using it wrong or something.
I always convert after every word, or it gets too frustrating for me. So, type word, press space repeatedly to get the kanji+okurigana I want, type some more, repeat. For strings of hiragana I type for as long as there won’t be any need for conversion, press enter, continue. It gets quite fast and seamless after a while.
At one point I used googles ime extension, iirc that broke down your sentence into words and allowed you to convert one by one. That’s a feature I miss dearly
I usually convert after a full sentence, but I think IPad IME converts to kanji as I go. If I need to find a less common word, then I will use more common readings of the kanji to build the Jukugo.
Trying to correctly convert every word in a sentence after the fact seems slower and more prone to error to me. Which is why I settled on the convert-after-every-word method. This way I’m also not training IME to expect common patterns from what I’m writing (because there are none!).
It was (nearly exactly) a year ago I too was so frustrated with the IME that I had to ask the forums: Typing frustrations
fcitx5 isn’t the IME itself. It’s the engine.
Probably in this case, Google-based mozc-ut is the best option. You might need to tweak a little though.
The IME you already use may already be tweak-able, like turning off suggestions based on history.
ah, right, I use mozc-ut for it
I use the same thing - are you going off the suggestions it gives you before you press space, by any chance? for instance when I type ひと the suggestions are 人が and 人が大嫌い (for some reason), but when I press space the top result is 人, followed by ひと, ヒト and 日と.
Pressing space should always limit the list to what you actually typed without any suggestions tacked on.
Ngl, I don’t really use space, rather I tab through the suggestions and press enter, is that the issue then?
Probably, yes - if you’re not using space you’re tabbing through a list of suggestions of what you might be wanting to type, which IMEs tend to be really bad at in my experience except for some often-used set phrases.
:o i do this also oops! thanks for the info Yami
I normally use Space before Up / Down, unless I want to use what the IME directly suggests. (Then, I would immediately press Up / Down).
This is as far as PC / laptop goes. I have no idea how to limit suggestions on Android (using Google keyboard right now, used to be SwiftKey.)
The arrow keys are too far away and are in an awkward place, that’s why I use Tab/Shift Tab, accomplishes the same thing tho
you should be able to hit space repeatedly and it will scroll through the list
at least that’s what it does for IBUS + mozc