Tempnet stress test_

Each tile is a real campaign transaction; larger tiles represent more serialized bytes on a logarithmic scale. New acknowledgements enter from the left, and observed confirmations move to the right. The first view places existing history without replaying it. This is a bounded sample of the 120 recent acknowledged transactions and40 recent confirmations, updated about every5seconds. Animation shows observed state changes, not measured network propagation. Quiet periods produce no artificial transactions.

Local block receipt intervals, sampled every two seconds, against the 60-second target. Observation gaps and catch-up batches are excluded. Header timestamp intervals are a separate series; individual blocks need not arrive on target.

Actual block size and consensus limit, with control and elastic components normalized to the one-minute interval. Each transition is checked against the pinned BCHN ABLA reference implementation. Sizes are decimal kB.

Confirmed-block fee rates in sat/byte: byte-weighted 10th, 50th and 90th percentiles, excluding coinbase. Coinbase-only blocks have no fee-rate point. Mined fees are not a guaranteed threshold for the next block.

Current campaign probes pay 1 sat/byte. The historical comparison retains earlier fee bands. Included ÷ eligible probes is shown for each cohort. Eligibility requires confirmed inputs, a stable submission tip, and acknowledgement at least ten seconds before the next canonical block is observed. Boundary races, observation gaps and dependent transactions are excluded. Counts matter: early percentages can represent very small samples.

These observations inform zero-confirmation assumptions. Size, relay, template timing and other miners also affect inclusion; neither confirmation nor DSProof coverage guarantees a zero-confirmation payment.