What Changed from the Original

This project is a fork and reimplementation of Androz2091's Discord Data Package Explorer. While we share the same core goal — helping users understand and explore their Discord data package — we have made several considerable changes.

High-level Differences

Area	Original	This Static Edition
Build System	Rollup + Svelte 3	Vite + Svelte 4
Architecture	Multiple views + routing + stores	Single-file reactive app (mostly in App.svelte)
Primary Focus	Statistics & insights (Wrapped-like)	Cross-server message search
External Dependencies	axios, diswho API, charts, routing	None (pure client-side)
Network Activity	Optional external user resolution	Zero network calls during processing
Message Content	Stripped for stats only	Full content kept for search
Deployment	Self-host or hosted instances	GitHub Pages (pure static, no backend)
Documentation	Basic README	Multi-language VitePress docs

Notable Changes & Additions

• Search as a first-class feature
  The original focused on aggregate statistics. We added powerful full-text search across all servers and DMs, with filters (servers only / DMs only), channel filtering, sorting, and CSV export. This directly addresses the common user need of "finding my old messages across servers."

• Complete removal of external services
  All calls to external APIs (for resolving usernames/avatars) have been removed. This makes the tool truly zero-telemetry and suitable for the most privacy-conscious users.

• Simpler, more maintainable codebase
  We consolidated logic into a single reactive component. This reduces complexity for a tool whose main job is "parse once, explore locally."

• Modern build toolchain
  Switched to Vite for faster development and simpler static export configuration.

• Strong emphasis on transparency
  The entire value proposition is "this is a pure static build on GitHub Pages." We added visible messaging, clear licensing for docs, and separate multi-language documentation.

• Documentation
  Added a proper documentation site (this site) with guides, architecture explanations, and privacy details. Initially available in English and German, with structure prepared for more languages.

What We Kept

• The excellent streaming ZIP parsing approach using fflate
• Support for both old (CSV) and newer (JSON) Discord data package formats
• Core statistics (message counts, hourly activity, top words, etc.)
• Demo data mode for testing without a real package
• Overall "everything runs in the browser" philosophy

Why These Changes?

The original project is excellent. However, our specific goal was to create a high-trust, purely static version that can be hosted on GitHub Pages with zero risk of hidden telemetry or backend processing. This required:

• Eliminating every possible external dependency
• Simplifying the architecture for long-term maintainability as a static artifact
• Prioritizing the search use case that many users actually want

We believe these changes make the tool more aligned with "pure static + maximum transparency" while still delivering real value.