掲示板 Forums - Hands free renshuu app

I love renshuu and I'd love to be able to use it passively too when I'm walking or driving. When I noticed there was an API, I thought I could make an hands free android app of guessing the word before it tells you the answer. In the same manner of some vocabulary test videos on youtube.

So far I've built the most barebones version with control of schedules and intervals.

I first only developped the recognition (japanese->english) and realized that was useless, way too easy. And was kind of frustrated to have spent weeks on the app for this. As a last resort I added the recollection option which is already harder, especially with sentences.

I wanted to get y'all opinion on this. And know If it would help other people than me. I hope I'm in the right topic.

Here is a list of features I could do :

• offline, never directly call the API, except for syncing data, so you can walk anywhere
• Fully passive mode, Like described. (only mode done)
• Interactive mode (lets you invalidate your answer with a gesture not needing screen focus)
  • invalidate using pause button on earphones
  • shaking the phone.
  • pause after reaveal and validate/invalidate and unpause with + and -
  • button on screen.
• Interactivity would make it possible to keep hard terms and make them appear more often on top of renshuu's vector data.
• being able to restrict the definitions
  • restrict meaning count, eg. Akai => 1. red,crimson etc.. 2.communist. being able to show only (few) first. here: 1. red,crimson.
  • restrict terms for each meaning separated by comma. can be long sometimes eg. Akai =>1. red, crimson, scarlet, vermilion. Only show red,crimson for example.
• experimental automatic pause duration depending on the length of the expected answer.
• sorting modes (harder first, shuffle, recent first etc..)
  • should implement a "today history", cause it currently start from the beginning if you stop. it would be cool to be able to start from when you stopped. (currently in harder first only)
• Differenciate recognition and recollection mistakes. currently i'm using the global userdata from the api for both, which is not accurate for recollection, i should use the vectors meaning ->japanese and japanese -> meaning errors instead respectively.

for now the things that prevent me the most from using it is not having enough time for long sentences and starting by the same words every time. so I think these are the main features to develop before even starting the interactive mode.
I also need to polish what I already have cause I currently dont have a stop button. and i want to have a view of only the text and the stop button appear on top of the view to select and start.

Little update.
I haven't worked on it during these 4 months. due to lack of time and motivation The days where I'd work on it I'd completely break my renshuu streak cause I can't seem to be able to do 50/50. I'm always 100% on one thing in a day. especially on activities that necessitate focus.
I do not intend to pursue the development for now. as it was mainly driven by a personal need that I no longer have as I now believe immersing in actual content is more important than just doubling SRS with audio. sure immersing's not a thing I'd do while walking (except if I listened to podcasts) but the hands free app would be adding cognitive load in a day on top of standard SRS and other types of immersion. as well as not really actually clearing reviews. just reinforcement.

I still do like the idea of using passive unused time to reinforce the terms / sentences. so maybe in a spark of motivation I'll work on it someday but that's unlikely.
here is the github for the project in case someone wants to clone or fork it: https://github.com/SHIRO-Suit/...
I don't know how to manage that, but If someone is interested we'll figure it out.
The code is far from clean though.