MIYO

Documentation

Public notes for users, release maintainers, and contributors.

This page behaves like a small README: concise, linkable, and focused on what the site and app actually do.

What MIYO is

MIYO is an Android comic and manga reader built for large libraries, online sources, and reliable offline reading. It is open-source and currently distributed through GitHub Releases.

Design priorities

  • Downloaded chapters should work without unnecessary network dependency.
  • Source-specific behavior belongs behind parser/plugin boundaries.
  • Long-running downloads need visible state, cancellation paths, and validation.
  • Native code should be measurable and recoverable, not decorative.

What this website does not do

This website does not host manga, mirror provider catalogs, bypass source restrictions, or replace the app's parser/plugin system.

Application layers

Keep the moving parts named.

The app is easier to improve when reader, download, source, native, and release concerns stay separated.

  1. UIhistory, favorites, explore, feed, local storage, reader, downloads, settings
  2. Readerchapter pages, cache, page state, reading progress, offline fallback
  3. Downloadsworkers, notifications, local directories, cancellation, validation
  4. Sourcesparsers, settings, authentication, captcha handoff, Cloudflare/browser flows
  5. NativeC++ helpers for image probing, archive work, and refinement acceleration
  6. ReleaseGitHub Actions builds, APK assets, model packages, update metadata

Compatible plugin direction

The current external plugin path is for Kotatsu-compatible JAR parser plugins. GitHub release imports should find JAR assets and load them through the parser bridge.

  • Use JAR plugins built for the MIYO/Kotatsu parser API.
  • Do not silently install unsupported APK extension formats.

Keiyoushi and Mihon APKs

Keiyoushi repositories are useful source ecosystems, but their APK extensions need a dedicated compatibility runtime before MIYO can run them safely. Until then, the app should explain the limitation instead of pretending the import worked.

GitHub Pages website

The website is static HTML/CSS/JS served by GitHub Pages from the main branch root. It intentionally has no build step.

Repository: Torro101/koharu-miyo
Source:     main branch
Folder:     /
Default:    https://torro101.github.io/koharu-miyo/

Custom domain note

No custom domain is configured yet. Use the default GitHub Pages URL until a domain is owned and DNS is ready.

Local preview

python3 -m http.server 4177
# open http://127.0.0.1:4177/