qalarc.com / projects / vk-fingerprint
VK Fingerprint
Privacy-preserving browser identity module — cookieless anonymous recognition via SHA-256 fingerprint from fonts, GPU canvas, hardware signals, and demographic anchor.
net — a platform where users post anonymously with demographic labels but no accounts. com, cached 24h), and demographic label (ethnicity:bracket:gender from the Folkus face scan). The raw signals never leave the browser — only the hash is sent to Cloudflare KV for returning-user recognition. Tokens expire after 90 days and auto-renew on each visit.
Includes a full test harness showing every signal live, token stability comparison (run A vs B), returning user simulation, and server registration tests.
Capabilities
Font enumeration — probes 30+ system fonts, detects presence/absence via canvas text width
Canvas GPU fingerprint — renders gradient + emoji + geometry, captures GPU/driver subpixel differences
Hardware signals — screen res, DPI, CPU cores, RAM, touch points, orientation type
Browser signals — timezone, language list, platform, vendor, feature flag support
GeoIP country — ip-api.com lookup, 24h local cache, flag emoji generation
Demographic anchor — ethnicity:bracket:gender locks token to person not just device
SHA-256 via SubtleCrypto — non-reversible, no PII in output hash
Cloudflare KV persistence — 90-day TTL, auto-renewed on each visit
Returning user recognition — localStorage check before any camera/server call
Live test harness — see every signal, compare runs, test stability, simulate returning user
Offline-safe — all API calls gracefully fail, local state always maintained
Identity twist support — stores true demographic + claimed override separately
Tags
Status: active · First built: 2026-04-22 · Last updated: 2026-04-22