Litecoin.watch beginner wallet and payment experiments, release 2026-10-03-v1.
Released 3 October 2026 (Europe/Warsaw). Test recorded at 2026-10-02T23:44:46.568380+00:00 UTC.
Core /LitecoinCore:0.21.5.8/; tested binary SHA-256 63fb440cdc5e2e7d46144b8e3a47dab4725fa9c056a77b8b12b11854504f4838.
11 wallet-state records; 24 payment constructions: 18 complete signed candidates accepted by local regtest policy, six insufficient-funds failures, one selected sweep broadcast and mined. 103 executable collector assertions passed.
Network isolated: new temporary regtest directory, listen=0, connect=0, DNS discovery disabled, zero peers. No mainnet funds or existing data directory. Only transparent native SegWit P2WPKH receipts; the mining reward is an ordinary test-chain coinbase. No mobile/GUI/hardware interface, MWEB, exchange account, real purchase, fee forecast or propagation benchmark.

COLLECT NEW TEST DATA
Save test-wallet-payments.py.txt as test-wallet-payments.py and isolated_rpc.py.txt as isolated_rpc.py in one directory. Authenticate your own Litecoin Core 0.21.5.8 binary; the recorded hash identifies our Linux binary, not every platform build.
python test-wallet-payments.py /absolute/path/to/litecoind /absolute/path/to/new-output-directory
Python standard library only. A temporary test node is stopped and its directory removed when the script exits normally. No private keys are exported. The public test-only encryption passphrase in the script has no relation to production credentials. Random keys mean identifiers differ on reruns; signature lengths can affect exact sizes. Re-run behavioral assertions and analyze your own result rather than requiring identical txids or every historical vsize.

RECALCULATE THIS FROZEN EXPORT OFFLINE
Save analyze-results.py.txt as analyze-results.py, then:
python analyze-results.py results.json
The byte reader parses transparent SegWit transaction inputs, integer output values and witness lengths, reconstructs stripped serialization, and recomputes txid, weight, vsize and input-minus-output fee. It matches declared input values to the exported outpoints. It does NOT cryptographically verify signatures, prove those inputs existed, validate consensus, prove wallet ownership or act as an independent peer. The local policy and mined outcome are observations recorded by our experiment.

UNITS AND INSTRUCTIONS
Money fields in CSV are integer litoshi (100,000,000 per LTC). Input groups each total 1 LTC: one 1 LTC output, three 0.33333333/0.33333333/0.33333334 outputs, or ten 0.1 LTC outputs. We explicitly selected every input in each group and disabled additional input selection.
Each group tests exact-full-balance, exact-0.4, subtract-from-0.4 and send-max at 2 and 10 litoshi/vB. fundrawtransaction feeRate uses LTC/kvB: 0.00002 or 0.00010. Do not copy those values into fee_rate, which has different units. The declared rates are experimental settings, not current recommendations.
Every candidate reuses its group's unspent funding outputs; summing all candidate debits would double-count alternative constructions. Only the selected one-input, 2-litoshi/vB send-max transaction was broadcast and mined after all candidates were checked.

BALANCE STATES
getbalance defaults can include watch-only value when private keys are disabled. getbalances separates mine/watchonly and trusted/untrusted_pending/immature. The selection-locked wallet retains its 1 LTC trusted balance but offers no automatic funding input. The encrypted wallet can construct a funded candidate while locked, but signing fails until local authorization is unlocked. These are different locks.
The ordinary incoming external receipt was excluded from default funding at zero confirmations and included after one confirmation; this is not a rule that all unconfirmed change is unusable. The isolated coinbase remained wallet-immature at 100 recorded confirmations and became wallet-available at 101. Litecoin's consensus constant is 100; Core's wallet calculation includes the additional depth offset. No elapsed-time promise is inferred.

RELEASE CONTENTS
results.json contains categories, non-secret wallet metadata, selected inputs, signed transaction bytes, decoded outputs, errors and assertions. The two CSVs are readable projections. Blank local TXT records collect no recovery secrets. SVGs show observed accounting or explanatory diagnostic steps, not real application screenshots. source-records.json identifies pinned source URLs and capture hashes; full third-party sources are not redistributed.
SHA256SUMS.txt covers all package content other than itself and the ZIP. research-package.zip includes that checksum file. The archive's separate checksum is in archive-sha256.txt to avoid circular hashing.
AI-assisted research, code and drafting; original vector artwork. No independent external reviewer or network-wide generalization claimed. Corrections: https://litecoin.watch/corrections/ . Original scripts, diagrams and blank worksheets: MIT; original results: CC0-1.0. Third-party marks/code retain their owners' rights.
