I have no idea, to be honest. Usually this is caused by separate UTF code points being written than what WK has in its system.
So I can debug this, please paste in the text you entered as well as the expected message text here or to howdy@kamesame.com. Be sure you’re typing the word using a Japanese keyboard and not copying and pasting something like a radical or a glyph from another keyboard type (e.g. Chinese)
Right, so I typed in 「うえ」hit space for 「上」, should be correct right?
Hi!
First of all, it seems like a great idea, I also always find myself to learn much better when I need to reproduce things in the target language.
I started using your application and then, a mis-click happened and I regenerated my WaniKani API v2 code. Now the app, expectedly, doesn’t sync to my account. It doesn’t give me any error though, is that expected?
Another thing is that I don’t know how how to change the API key. Is that possible? What should I do now?
Thanks
Should be but I need the actual characters as text to verify the code points differ (or don’t)
You plan to add audio support when it is done in API v2?
I didn’t even know it was being added!
It only says “in fact, yadda yadda” if there’s an exact match in WK
Have you considered the option of marking right answers wrong? I just wrote OK a kanji but the reading I was thinking was totally off, seeing it marked as OK feels like cheating 
Oh you mean like a button to revoke your own credit from the summary page?
Hmm. Interesting idea would be a lot of work though.
Ummm, dunno if I’m explaining myself wrong or understanding your reply wrong -I’m not a native english speaker-. What happened to me is that I typed the right answer correctly using kanji, but when I submitted it I saw the reading I was thinking was off. So, when I hit the button to go to the next review, that review goes to the group of “OK” ones. It would be nice to have a button to make if go to the group of KO ones so you get it asked again in the review series.
Thank you so much for making this! Finding it so so useful
UTF stuff:
囗 - U+56D7 - is marked as wrong for Mouth. KameSame expected 口 - U+53E3 -
the 56D7 codepoint is the box radical, not a kanji. How did you manage to type that?
I have a sneaking suspicion that some number of users with problems like these are copy-pasting characters out of WK or some other source rather than using their OS’s built-in Japanese keyboards. There’s nothing invalid about the former (as long as the character matches), but the intent of the app is to give folks practice typing on a normal Japanese keyboard / IME
I’m using Google’s handwriting keyboard for japanese. So your suspicion is totally right. But I think apps like KameSame or KaniWani + handwriting keyboard are a great practice.
Haven’t taken a screenshot (will try when I get this review again) but I got question about “long time”, to which correct answer is 久しい. I entered just 久, and then I got this message:
“Although 久 does mean “Raptor Cage”, we were actually looking for 久しい”.
“Raptor cage” is name of radical, but I think that you should rather get description from kanji…
In that case I’m not really sure what to do, since radicals are indeed separate characters. It seems like the Google handwriting IME is the one in the wrong here.
Thanks! This should now be fixed
I’ve been thinking about it and probably it is up to us user to get used to the annoyances of those keyboards.
If there was an easy way to hack in support for what you’re trying to do, I would, but there are too many radicals, and a lot of radicals deviate sufficiently from their single-radical kanji that I’m wary of it