Litecoin.watch transaction lab — recorded 26 September 2026

Independent verification (Python standard library, no network):
  python verify-evidence.py results.json
Reproduction, using a verified Litecoin Core 0.21.5.8 executable:
  python run-experiment.py /path/to/litecoind new-results
On Windows supply the quoted full path to litecoind.exe.
Rename downloaded *.py.txt files to *.py first. Do not run with python -O.

The experiment creates fresh temporary regtest directories, cookie-authenticated
localhost RPC and test wallets. Peer listening, outbound connections, DNS seeds
and discovery are disabled. Both child nodes stop on completion. Temporary test
directories are retained. No existing wallet or mainnet directory is opened.

12 signed fee-size candidates and 3 payment amounts are measured. One native
SegWit payment is broadcast on regtest, loses its confirming branch, returns to
the mempool and is mined again. Blocks are withheld and manually delivered over
RPC: this does not measure internet propagation or actual block waiting times.
Both outputs use P2WPKH. No MWEB, Taproot or multisig size claim.

Requested funding rate: 0.00002 LTC/kvB = 2 litoshi/vB. Actual fees and signed
vsize are separate measurements: fee estimation precedes final signing. A new
run generates different keys, signatures, IDs and times, and may differ slightly
in size. Reproduce the relationships and invariants, not byte-identical IDs.

results.json includes signed raw candidates, previous-output amounts, measured
sizes, node-state observations and assertions. The independent verifier checks
decoding, IDs, weight and conservation; it does not independently verify ECDSA
signatures or prove the node-state transcript. Core performed signing and
policy checks during this controlled run.

Data and original scripts: CC BY 4.0, attribution to Litecoin.watch.
Litecoin Core remains under its own upstream licence.
https://litecoin.watch/transaction-lab/