I am considering writing a Userscript for WK that addresses the biggest obstacle I face when learning vocab: kanji readings.
Specifically, when I learn a new kanji (so not vocab yet), I learn exactly one reading, which is sometimes onyomi and sometimes kunyomi. But there is nothing in the system that drills me on knowing which is which until the vocab comes along. For a bunch of my kanji I have successfully Guru’ed or Master’ed, I can give you both, but for the life of me I couldn’t tell you which is on and which is kun. Sometimes I can reverse engineer from vocab: “Well, I know readings A and B, and the kanji occurs in jukugo X with reading B, so that means B is probably the on, and A is the kun.”
This is annoying, and I am thinking that, for me personally, it would probably be better if the kanji reviews would drill me on 3 items instead of 2. Now it’s the meaning and “the reading” which may be on or kun, but instead it would drill me on the meaning, the primary onyomi and the primary kunyomi, and the review would only count as correct if I get all three right.
I tried this out for a bit with an experimental Anki deck generated via the API, and it seems to work quite well. (And no, I’m not sharing that deck. Partly because it’s rough and experimental, and partly because I want to stay on the good side of WK’s terms of service.)
So I’m considering turning this into a Userscript since I don’t want to give up on WK’s smooth spoon-fed SRS. Is this something others might also be interested in? No promises, I just want to see if I should consider making this a public thing for others to enjoy as well.