“It’s easy to do for kanji, since we just pop in the other reading, but it wouldn’t really be a practical solution for reading typos since we’d have to think up and then manually define thousands of instances.”
Thanks for your feedback! Although, I wasn’t talking about typos for the meaning. What I am talking about is that a lot of times there will be several vocabulary prompts in a row that ask for the reading of the words, and then suddenly one will appear which asks for the meaning. Because I just entered the reading of the past few words, my brain automatically thinks I should input the reading and not the meaning. For example, yesterday I saw the word “前” and I tried to type in まえ, but it was counted as wrong (understandably so) because it was looking for “front.” It is a small grievance, but if it shook and said “that’s the reading. We’re looking for the meaning instead,” it could lead to a lot less frustration when working. Also, while there are a lot of vocab words, there are only two main methods that are used for romanization of Japanese words, so inputting those (in this instance “Mae”) wouldn’t be too much of a task (as opposed to inputting typos of the correct answers, which would be impossible). Again, though, thanks for your feedback!