Skip to contentQbit Wallet is live — add it to Chrome
← All releases
CoreTestnet4

Qbit Core v0.1.2-testnet4

Release notes

Qbit version v0.1.2-testnet4 is available from the release page for that version.

This is a public testnet4 maintenance release based on the changes since v0.1.1-testnet4. It preserves the reset testnet4 lineage from v0.1.0 and focuses on AuxPoW mining compatibility, P2MR wallet and Qt signing behavior, witness-pruned startup safety, P2MR/PQC script and data-signature coverage, release/CI hardening, and public testnet operator documentation.

Testnet-era release artifacts are for public testnet use only unless the release page explicitly says otherwise. Qbit mainnet is not launched. Do not treat any in-tree mainnet genesis block, hash, seed placeholder, or endpoint placeholder as an official mainnet launch commitment.

Testnet coins have no economic value, and the public testnet may be reset or replaced during rehearsal.

Please report bugs using the Qbit issue tracker.

How to Upgrade

Shut down the older qbit process first and wait until it exits completely. Then install the new binaries from the GitHub Release page.

This release does not require a testnet4 chain reset for nodes upgrading from v0.1.1-testnet4.

On Linux, replace the existing qbitd, qbit-cli, qbit-tx, qbit-util, and qbit-wallet binaries with the new versions.

On macOS, replace the installed Qbit application bundle or unpack the updated archive into the desired location.

On Windows, run the installer if one is provided for the release, or replace the unpacked binaries with the new release artifacts.

Compatibility

The public Qbit testnet4 chain is selected with:

qbitd -testnet4
qbit-cli -testnet4 getblockchaininfo

Official testnet release binaries are expected to reject no-flag mainnet startup and explicit -chain=main startup. Use -testnet4 or -chain=testnet4 for the public testnet.

Supported and tested platforms for this release are the artifacts attached to the GitHub Release page. The release page is the source of truth for available platform builds, checksums, signatures, and attestations.

Public testnet4 network identity remains the v0.1.0-testnet4 reset testnet4 identity unless the release page explicitly says otherwise.

There are no scheduled consensus activation changes in v0.1.2-testnet4. Mining pools, exchanges, explorers, wallets, and other testnet4 operators can upgrade without a chain reset.

Item Value
Chain flag -testnet4 or -chain=testnet4
Default P2P port 48355
Default RPC/REST port 48352
Address HRP tq
AuxPoW chain ID 31430

Notable changes

Consensus, mining, and node operation
  • submitauxblock now accepts legacy Dogecoin-style AuxPoW payloads that include the extra parent-block hash field, normalizes them into qbit's canonical AuxPoW layout, and validates/stores the canonical form.
  • Permissionless getblocktemplate responses now sanitize regtest -blockversion overrides that accidentally set the AuxPoW signalling flag. Permissionless templates remain canonical with chain_id=0, auxpow=false, and preserved low versionbits.
  • The mining pool quickstart now documents the commitmentorder and commitmentactivationheight fields returned by createauxblock, and tells pool software to follow the node-returned commitment order instead of Namecoin, Dogecoin, or serializer assumptions.
  • Witness-pruned -reindex-chainstate startup now validates the supplied -assumevalid hash before any local chainstate or snapshot chainstate wipe can run. Null, unknown, off-best-header-chain, insufficient-work, and too-low hashes fail before destructive startup paths.
Wallet, RPC, and PQC signing
  • Qt Sign/Verify is enabled again on P2MR-only chains, with P2MR data-proof labels instead of legacy message-signing labels.
  • Qt Sign/Verify can sign UTF-8 text after hashing it, or sign raw 32-byte hashes directly, with wallet-owned P2MR/PQC keys. It emits proof JSON compatible with signdatapqchash.
  • Qt Sign/Verify can verify pasted P2MR/PQC proof JSON compatible with verifydatapqchash.
  • Qt signing locally verifies generated proofs before displaying them and shows PQC signature-budget state after signing.
  • P2MR data-proof verification is shared between the Qt verifier and the verifydatapqchash RPC path.
  • Wallet-backed PQC data-hash signing now treats persisted counter reservations as consumed once the reservation succeeds, even if the later raw signing operation fails.
  • signdatapqchash reports and logs consumed PQC counter ranges for data-hash signing attempts that fail after a reservation has already been persisted.
  • P2MR data-hash signing now distinguishes exhausted PQC signature budgets from missing private keys when a selected leaf cannot be signed.
  • Provider, wallet unit, and functional RPC coverage now exercise successful and failing PQC data-hash signing counter paths.
  • New qbit-qt wallets now default to encrypted creation for ordinary private-key wallets. The Create Wallet dialog labels encryption as recommended and explains that encryption protects private keys while enabling authenticated wallet loads that avoid repeated plaintext PQC key validation.
  • P2MR getnewaddress and MarkUnusedAddresses flows avoid full synchronous keypool top-ups when ranged P2MR descriptors remain above the low watermark. They schedule bounded low-watermark refills while preserving exhaustion, locked-wallet, recovery, and rescan behavior.
  • Explicit wallet output types must now be both chain-allowed and backed by a matching wallet manager on P2MR-only chains.
  • fundrawtransaction now rejects prebuilt raw outputs that are not allowed in restricted-output mode before wallet funding. P2MR, OP_RETURN, and PayToAnchor outputs remain accepted where policy permits them.
  • deriveaddresses RPC help now labels rawmr(<merkle root>) as an expert-only arbitrary-root descriptor form whose roots cannot be proved or signed by the wallet.
P2MR descriptors and future leaves
  • Added an end-to-end feature_p2mr.py regression for rawmr(<future-leaf-root>) outputs.
  • Covered public descriptor derivation, wallet funding, signer refusal for a non-0xc0 future leaf, standard mempool rejection, and mandatory block acceptance for the same reveal spend.
  • Added focused unit and functional coverage around reserved P2MR leaf-version policy, masked future leaf control bytes, and PSBT signing/finalization refusal for reserved leaf versions.
  • Added independent JSON vectors for P2MR leaf hashing, branch hashing, control blocks, Merkle roots, scriptPubKeys, and mainnet/regtest addresses.
  • P2MR commitment-vector tests now cover malformed control blocks, wrong leaf versions, wrong CompactSize or script-length commitments, mutated siblings, unsorted branch roots, and wrong root commitments.
  • Descriptor, RPC, script, and PSBT tests now reuse the independent P2MR commitment vectors.
P2MR script, sighash, and data-signature coverage
  • Added fixed byte-vector P2MR script-path tests for a single-leaf spend and a two-leaf branch spend.
  • Added a fixed byte-vector CTV spend test that avoids deriving the accepted spend from production commitment helpers.
  • Added independent P2MR/PQC witness-vector coverage for the default OP_CHECKSIGPQC spend path, including mutated signatures, wrong sighash domains, bad hashtype suffixes, and wrong script/pubkey commitments.
  • Added an independent annex-present P2MR OP_CHECKSIGPQC witness vector with annex spend-type encoding, annex hash commitment, no-annex contrast data, and annex-specific near-miss failures.
  • Added independent fixed P2MR OP_CHECKSIGPQC witness vectors for non-default sighash modes, including SIGHASH_NONE, SIGHASH_SINGLE, and ANYONECANPAY combinations.
  • The fixed P2MR PQC witness fixture now pins serialized spend transactions, spent outputs, P2MR leaf/control data, public keys, signatures, raw p2mrSigMsg data, and final p2mrSighash values.
  • Added reproducible generator coverage for the P2MR PQC witness vectors so the checked-in fixtures can be regenerated independently.
  • Added codeseparator-specific P2MR OP_CHECKSIGPQC vectors for default and non-default OP_CODESEPARATOR positions, including same-leaf branches with different codeseparator positions.
  • Added negative coverage for wrong codeseparator positions, domains, public keys, scripts, and signatures.
  • Added explicit SIGHASH_SINGLE P2MR cases covering both matched-output acceptance and missing-output rejection.
  • P2MR SIGHASH_SINGLE tests now assert failure for missing outputs instead of allowing the legacy uint256::ONE behavior.
  • Added independent OP_CHECKDATASIGPQC P2MR witness-vector coverage for accepted data-signature spends and near misses involving corrupted signatures, wrong domains, mutated witness message hashes, mutated leaves, and wrong public keys.
  • Added independent OP_CHECKDATASIGADDPQC P2MR witness-vector coverage for threshold data signatures, including n-of-n success, m-of-n success with empty-signature skips, threshold failure, invalid non-empty signatures, wrong message hashes, wrong committed keys, wrong domains, and mutated leaves.
  • Added pinned P2MR data-hash proof fixtures for the QbitDataSigPQC domain, including C++ fixture validation, wallet signing reproduction, and verifydatapqchash functional coverage for malformed or mutated proof data.
Release and CI
  • Release publication validation now requires the signed release tag target to be an ancestor of the configured trusted_release_ref.
  • Public release-trust documentation records the v0.1.1-testnet4 signed tag object and target from public git state.
  • Public branch rulesets now use a path-aware Required Merge Gate profile so source, release-policy, RPC docs, public docs, and GitHub metadata changes receive appropriate validation.
  • Scheduled/manual RPC performance, IBD performance, and lint image prewarm workflows now target the public 0.1.x profile.
  • CI Docker BuildKit setup is covered by a tested helper, preserves the internal registry configuration where needed, and adds resolver fallback for public package hosts.
  • CI profiles are portable across public and internal repository contexts, including hosted/public images and trusted-runner paths where appropriate.
  • Nightly CI and RPC performance workflows were hardened by pinning scanner tooling for the available Rust version, separating qbit-only test-node capabilities from Bitcoin Core reference nodes, and sharding long ARM unit tests.
  • P2MR-heavy wallet unit tests were split into smaller CTest-visible shards so expensive keypool, descriptor, encryption, validation, signing, and trusted-record coverage receives separate timeout budgets.
  • The libbitcoinpqc subtree verifier now checks that src/libbitcoinpqc matches the peeled upstream tag tree and recorded git-subtree-split, and the checker is wired into subtree lint and update tooling.
  • Added a focused qbit single-PR review skill with qbit-specific consensus, networking, validation-evidence, and GitHub-ready review-output guidance.

Known issues

No new release-specific known issues are added by v0.1.2-testnet4. Testnet4 remains a public rehearsal network and may still be reset or replaced during the testnet era.

Credits

Thanks to everyone who contributed code, testing, review, infrastructure, and release coordination for this release.