METHOD / ACTIVITY MODE
Real swaps.
No dollar bottleneck.
The instruments show swap activity—not estimated trade value. Real canonical Gateway swaps fuel eruptions, glow and recorder peaks even when USD cannot be trusted.
What powers the visuals?
Known source block times group real swaps into ten-second bins. Particle radius and recorder height increase logarithmically with the observed count in that bin, saturating at 32 swaps. Crater glow follows the last minute's observed swaps, saturating at 120.
A burst means at least twelve observed swaps in a ten-second bin. Streaks, count/10s labels and recorder wiggles are activity cues, not dollar-valued whale claims. Sampled swaps/min divides the timed sample by its covered minutes, with a one-minute minimum denominator. The sample is bounded and may be incomplete.
Motion and timing
Volcano replays up to 24 recent real swaps per venue once over about 6.5 seconds on arrival, explicitly labelled RECENT SWAP REPLAY. This is not newly arriving trade flow. New observations then spread over about 13.5 seconds. Particle scheduling is illustrative, not exact trade timing.
The recorder positions known-time events at their real block times. Unknown times get observation-span bands, not invented timestamps. Missing coverage is hatched. Empty ink never proves zero trading. Reduced motion keeps the same activity data and peaks but stops continuous animation; hidden tabs pause.
Keep the uncertainty.
The source is the canonical Avalanche Gateway, sampled at up to 500 swaps each for Pharaoh, LFJ and BlackHole per refresh. Browser history is bounded. The venues can cover different periods, so sampled counts/rates are not full-market rankings or 24-hour totals.
Unsupported USD remains absent from the API and is not used for any visual cue. Buy/sell stays unknown unless explicitly supported. No wallet identity, route or profit is invented. A venue outage says Unavailable, never zero; stale data keeps its original times and source warning.
Using the pages
Both views use the site's saved light/dark choice. Visible tabs refresh every 15 seconds with bounded retries; failed observations do not become fictional swaps. The startup replay is made only from recent real records and never replays a hidden-tab backlog.
This is an observational art instrument, not an execution terminal or trading recommendation.