wikipedia3d

Wikipedia3D (legacy revival)

A 2011-era WebGL + Semantic Web mashup that renders Wikipedia/DBpedia entities in a navigable 3D space, with map-based discovery and a rich UI for favourites, blacklisting, and exploration. This repo contains the original PHP/JS frontend and lightweight proxy scripts used for API calls.

---

Quick assessment (what you have today)

Strengths

• Unique concept still feels fresh: spatial exploration of knowledge, fast discovery loop (search → spawn 3D item → navigate → branch).
• Feature-rich UX: favourites, blacklist, keyboard navigation, skybox themes, map-driven discovery.
• Fully client-driven rendering with Three.js and simple PHP utilities—no heavy backend required.

Pain points / risks

• Very old JS stack (jQuery 1.6, MooTools, ancient Three.js) and global namespace design.
• Deprecated APIs: Google Maps “panoramio” library is discontinued; AlchemyAPI is defunct.
• Hard-coded domain URLs (wikipedia3d.world.eu.org) break local use and make the app fragile.
• Mixed-content/HTTP: many endpoints are http:// which modern browsers block on https://.
• CORS + proxy reliance: API calls are routed via PHP proxies; security + reliability needs care.

---

Quick local hosting (FastAPI replacement)

The legacy PHP proxy has been replaced by a small FastAPI app that also serves the static site.

cd /home/lucky/.openclaw/workspace/wikipedia3d
python3 -m venv .venv
source .venv/bin/activate
pip install -r server/requirements.txt
uvicorn server.app:app --reload --port 8088

Open:

• Landing page: http://localhost:8088/
• App itself: http://localhost:8088/wikipedia3d/

⚠️ For best results run over HTTP, not HTTPS, because the code uses legacy HTTP endpoints.

What changed

• server/app.py implements /proxy (replacement for ba-simple-proxy.php).
• wikipedia3d/index.html and /index.html are static HTML replacements for the old PHP pages.
• Proxy + icon URLs now point to relative paths to work locally.

---

Outlook: modernizing while keeping the soul

Goal: Preserve the core experience (3D exploration + semantic discovery + map-driven expansion) while upgrading to a modern, maintainable stack.

Phase 1 — Stabilize (lowest effort, highest payoff)

• Replace hard-coded URLs with relative paths.
• Add .env or config file for API endpoints/keys (Google Maps, GeoNames).
• Move inline scripts to files; reduce global pollution.
• Add a tiny local dev server script (Node or PHP).

Phase 2 — Modern frontend

• Upgrade to a current Three.js and refactor rendering to ES modules.
• Replace MooTools and legacy jQuery usage with vanilla JS or a minimal modern library.
• Use fetch() + async/await instead of jQuery AJAX.
• Replace the deprecated Google Maps bits (Panoramio) with Maps JS API + Places/StreetView or Mapbox.

Phase 3 — Data + semantics refresh

• Migrate from DBpedia-only to Wikidata SPARQL as primary, keep DBpedia as fallback.
• Replace AlchemyAPI (dead) with modern entity extraction or just remove it.
• Add caching and rate limiting in the proxy.

Phase 4 — Experience upgrade (keep the magic)

• Improved camera controls (orbit + FPS toggle).
• Smooth item layout + collision avoidance.
• Mini-map + bookmarks + “topic trails” visualization.
• Optional VR/WebXR mode while retaining keyboard-first navigation.

---

Key dependencies (legacy)

• jQuery 1.6.2, jQuery UI 1.8
• MooTools 1.3
• Three.js (very old build)
• Google Maps JS API (panoramio library deprecated)
• DBpedia SPARQL, Wikipedia API, GeoNames
• PHP proxy scripts (ba-simple-proxy.php)

---

Next steps (suggested)

1. Confirm the app runs locally with the PHP server.
2. Apply the URL fixes above.
3. Decide on a modern map provider (Google Maps w/ key vs Mapbox).
4. I can then create a clean “modernization branch” with ES modules + updated Three.js while preserving the UI flow.

---

If you want, I can start a staged refactor plan and implement Phase 1 and Phase 2 step-by-step.