KameSame - a fast, feature-rich Japanese memorization webapp

Would love to take a look at the code. I’m thinking it’s a private repository though, since I couldn’t find it on your Github profile?

Do you mean more times than one to confirm the JP input and then one more to submit the answer? Like literally more than a couple? Because two is normal.

Yeah, this is definitely high on the list (pulling only lessons at whatever WK progress level you want)

That’s interesting. It’s possibly just server lag, as we do a roundtrip HTTP PATCH to the server for every answer provided (since we’re taking your text to determine which alternate matches you might be running int)

Yeah, I think I’m going to plan on keeping this closed source. My motivation is a lot higher with this being a dogfooding personal project I care about than to have a bunch of other eyes on the code & backlog with me having (yet another) OSS repo in a maintainer role.

I’m not at all opposed to getting more help from others, but I think I want to give this at least a few weeks of work to stabilize before doing so

Sorry, I’m new to the forums and didn’t realize PMs were disabled. Feel free to e-mail me howdy@kamesame.com with whatever diagnostic info you can muster (I’ll also be checking error reporting to see if it gets me any insight)

1 Like

This looks almost exactly like an ideal SRS for English-Japanese direction which I described in another thread some time ago. However, I believe it shouldn’t reveal the expected answer when the user gives a valid answer that just happens to not be what the app is looking for. It should just show the information that the answer is valid but the app is looking for something else, so that I have to properly use my memory to remember the synonym. Please at least consider making this configurable.

5 Likes

I love it! I don’t know if this has been suggested before, but I would like to review only vocabulary, so an option to omit radicals, kanji or vocab would be good. It’s not a big deal though, and I don’t know if many other people would like this approach so if it’s not implemented I will still be using your app. Thank you for taking the time to do this, it’s awesome! :durtle_cat:

1 Like

I appreciate this but I think once you get to higher levels, it’ll become overwhelming. Some of these words (esp when you’re loose on matching exact spelling) have upwards of 15-20 alternate matches, and that’s before I start pulling in synonyms. You could spend all day racking your brain on one word

The app already omits radicals, since they’re not pronounceable and many are not easily reproduced as UTF characters. I’ll consider this option but I don’t know how I feel about it, because remembering onyomi without the context of a broader word is important. Very often I’ll fail to remember the onyomi of a kanji by itself but remember it in the context of the half dozen other vocab I know, and that sets me up for failure later when I see it in a kanji compound I don’t already know

True, but isn’t that what you are practicing on WaniKani already anyway?

The vocab only aapproach of KaniWani always made sense to me. When going E=>J the only thing I would need in real life is vocablary.

1 Like

You do learn how to recognize the character & the onyomi in WK, but seeing an english word for a Kanji and producing the kanji isn’t the same thing but is separately valuable. It exercises your knowledge of the onyomi by forcing you to produce it with the keyboard and to visually discriminate it from other kanji in order to produce it. I definitely would not dismiss it as somehow less valuable than vocab.

Yes, you have to put in the Kanji for the site to recognise it

I just wanted to pop in and thank you for putting that site up, for listening to the feedback and all that good stuff! You rock :smiley:

1 Like

@searls any hope for adding the reading to successful results. Would help to avoid typing in the wrong reading, as both mobile and IME are really good in correcting obvious mistakes.
Similar to your above reasoning I can see that ‘wrong way of entering’ is not an error but aiming for consistency is helpful for long time memorization.

Ah this is a valid point and something I can easily add

Will there be a change to what gets imported?Guru’d and higher?

Yes, I’m working on making this configurable.

Blah. yes, we should do something about the various 〜 words. It is definitely typable on macOS & iOS but I don’t know if it’s possible on Android or just hidden.

1 Like

Looks interesting. I thought I’d give it a try, and while doing my first lesson batch, got this response: