Interrupted publication
One of four chunks produces an incomplete result, a nonzero exit and no partial plaintext file.
THE PROOF / MEASURED, REPRODUCIBLE
The publisher stops.
The file comes back.
A real recovery result is the foundation. Product ambitions and laboratory measurements have different jobs; here are the measurements.
EXPERIMENT 2 · LOCAL BITCOIN REGTEST
89,527bytes in.
Identical bytes out.
Identical bytes out.
3 authenticated recordsSeparate recovery nodePublisher offline
Recovery without
the original environment.
The integration test publishes a synthetic JPEG, stops the publisher and moves its data directory offline. A separate Bitcoin Core node supplies the historical blocks. Recovery runs from a fresh working directory with receiver wallets disabled.
The reader receives the media secret and a starting height, together with network access to the receiver. It is not given the publisher’s journal, original file, wallet or transaction list.
The reconstructed JPEG matches the original byte for byte. This verifies the mechanics in a controlled regtest environment; it does not measure public-network censorship resistance or camera security.
RECORDED 05 SEPTEMBER 2026 / VERSION 2
| Input | Original bytes | Records | Commit + reveal fees | Result |
|---|---|---|---|---|
| Synthetic JPEG | 89,527 | 3 | 25,451 test sats | Byte-identical |
| Binary fixture | 100,000 | 4 | 33,894 test sats | Byte-identical |
Local regtest measurements. Test sats are not a USD price or a Litecoin fee estimate. Older experiment-1 figures do not describe the current padded format.
BEYOND THE HAPPY PATH
Wrong secret
No recovered file. The result is not_found, as it is for an interval with no matching objects.
Chain reorganization
Removing the last reveal block makes recovery incomplete. Restoring it restores successful recovery.
Hostile records
Forged ciphertext under a copied public header is rejected and counted without poisoning valid recovery.
Keep the claim.
Extend the evidence.
The recorded public testnet4 work prepared and inspected signed transactions without broadcasting experimental media. The product direction now prioritizes Litecoin, with Bitcoin as additional redundancy.
The results on this page cover the documented Bitcoin reference tests. They do not establish a shipped camera, a completed Litecoin adapter or a public-chain recovery demonstration.
Read the complete test report ↗