Migration to Tauri

[Sytronik]

Motivation

1. Electron renderer process w/ node integration doesn't work with Web Workers well.
   • We need multi-threaded wav rendering.
2. Tauri is designed for Rust backend.
3. Tauri has small binary size & fast start-up time.

Caveats & To-dos

• Tauri uses system webviews. We need to handle OS-Webview dependency.
• Tauri's IPC functions are async. It makes React codes complicated.
  • It'd be better to move some backend functions to wasm module. (especially functions repetitively called in a short period)
  • We might need to refactor frontend by gathering most of the BackendAPI call to a new hook / context provider / ... to avoid triggering too many re-rendering by too many Promise.then(setSomeState).
• Tauri's IPC is slower than calling n-api function in the same thread.
  • In the main branch, backend returns sliced spectrogram. It is problematic with tauri's IPC w/ JSON serde.
    • We also need to move spectrogram mipmap code to wasm module.
• (Pros) Tauri backend can emit events to the frontend.
  • We can refactor some backend functions. e.g., get_player_state

This PR covers

• The same behaviour as the main branch w/ Tauri on macOS, Windows
• Ensuring smooth wav & spectrogram rendering
  • powered by multi-threading w/ Web Worker + WASM
• Merging main branch and fix-front-for-windows branch
  • We need to handle more OS compatibility issues with Tauri.
  • We can solve those issues while solving the issues that fix-front-for-windows branch handles at the same time.
• Avoiding other refactoring that doesn't change user's experience dramatically
  • e.g., gathering BackendAPI call to a new module, emitting events from backend to alert player state