Building JapanApp: From Vision to SRS Architecture

Learning Japanese Kanji can be overwhelming. With thousands of characters, stroke orders, Onyomi, and Kunyomi readings, rote memorization without systematic intervals leads to high drop-out rates. I decided to build JapanApp (japan-mee.vercel.app) to tackle this challenge directly.

The Technical Challenge The core requirement was twofold: 1. **Interactive Stroke Order Canvas**: Rendering SVG vectors dynamically and evaluating whether user finger or cursor strokes follow proper stroke trajectory without high latency. 2. **Spaced Repetition System (SRS)**: Implementing an algorithm inspired by SuperMemo SM-2 to predict optimal review intervals based on difficulty feedback.

Architecture with Next.js & TypeScript Using Next.js alongside React and TypeScript allowed for server-rendered kanji dictionaries while keeping the interactive flashcard practice strictly client-side for zero-lag responsiveness. By persisting progress with local state synchronization, learners can study seamlessly with rapid feedback loops.