Evaluating Deterministic BFT Finality within a Confidential UTXO State Machine

The architecture employs deterministic Byzantine Fault Tolerance (BFT) state transitions, avoiding traditional longest-chain consensus and related reorganization risks. Instead, the protocol finalizes blocks in epochs, each consisting of 32 blocks, with finality achieved as soon as a supermajority cryptographic quorum from the epoch's committee signs off on a block hash. Committee selection for each epoch is deterministic, based on validator stakes recorded in the finalized chain state, ensuring dynamic yet predictable participation in consensus.

Privacy in FinalisCore is implemented at the cryptographic level using secp256k1-zkp, similar to systems like Elements and Liquid. This enables validators to confirm transaction balance correctness through homomorphic addition, allowing them to verify that inputs and outputs match without revealing underlying asset amounts. Such an approach provides robust confidentiality for all transactions, maintaining privacy guarantees even as the protocol scales.

The project has reached a public testing phase, with testnet-1 deployed to assess liveness and safety under various network conditions, including intentional partitions. Formal specification and validation have been conducted using TLA+, and the team is now focusing on scaling beyond a single bootstrap validator toward a decentralized configuration involving multiple independent node operators. With this expansion, several technical challenges are being prioritized for community feedback, particularly around mempool invalidation during rapid committee shifts and the computational overhead of batch-verifying Pedersen commitments.

In the context of mempool policy, there is a clear separation between consensus rules and node-level relay thresholds. Committee rotation should not impact the fundamental validity of transactions; rather, wallets must be designed to re-evaluate their input selections against the updated, finalized UTXO state after each epoch transition. For optimizing proof throughput, it is recommended to independently benchmark transaction parsing, range-proof verification, and state update processes before applying cache optimizations. Additionally, when implementing batching mechanisms that utilize randomized coefficients, care must be taken to ensure that acceptance criteria remain strictly consensus-deterministic, never relying on node-local randomness.

All resources for review and contribution are available under the MIT license, including the complete repository, the detailed technical whitepaper, and setup instructions for the live testnet environment. All tokens used in the testnet are purely for experimental purposes and hold no real-world value. Community engagement is encouraged, with particular interest in architectural critiques, code reviews, and protocol-level feedback to further refine FinalisCore’s unique trade-offs and operational guarantees.