Manga/Comic Reading Platform

2024

Two related reading platforms with microservice backends, real-time progress tracking, image processing pipelines, and offline-capable PWA frontends.

3 backend microservices

Sub-200ms image serve times with Sharp processing

iOS-optimized memory management for image-heavy pages

The Challenge

Building a fast reading experience with image-heavy content, offline support, real-time download progress, and full-text search across large content libraries, while reverse-engineering undocumented APIs.

Approach

Built 3 backend microservices: a downloader with Socket.io for real-time progress, a streamer using Sharp for image processing, and an indexer combining Express with a Go binary for filesystem scanning. Frontend is a SvelteKit PWA with infinite scroll, IndexedDB offline storage, and iOS memory optimizations.

Key Highlights

Real-time download progress via Socket.io
Sharp-based CDN-like image batching and streaming
Go binary indexer for full-text search across filesystem
Offline-capable PWA with IndexedDB storage
Reverse-engineered undocumented API endpoints

Technologies

TypeScriptExpress 5SvelteKit 2Socket.ioSharpGoPM2IndexedDB