ADR-0001 — Blockchain framework and consensus: Hyperledger Besu (QBFT)
Status: Accepted Date: 2026-07-29 Decided by: Tamga Network project management
Context
PM-BC-0001 decided that Tamga would run its own sovereign, open-source, permissioned network with equal power per state and BFT/PoA consensus, but deliberately left the concrete framework/engine choice (layer 4) open.
RS-FRAMEWORKS-0001 compared two candidates:
- Hyperledger Besu (QBFT)
- Cosmos SDK (CometBFT)
They were assessed against Tamga's requirements: permissioned, equal votes, immediate finality, sovereignty, a single shared chain, future asset and payment capability, and interoperability with EBSI.
Decision
Tamga Network's blockchain engine is Hyperledger Besu, running QBFT consensus.
Sub-decisions:
- The network is set up as an independent, permissioned private network (its own genesis, its own chain ID). It is not connected to Ethereum mainnet or any other network.
- Consensus: QBFT (2/3+ supermajority, single-block finality). Validators have equal voting power (PoA).
- Validators = states; full nodes = trusted institutions (for scaling and the equal-power principle).
- Smart contracts: EVM / Solidity — future asset, transfer and payment capabilities are built on this layer.
- The final choice between QBFT and IBFT 2.0 is settled during implementation (this ADR may be updated).
Rationale
- Native EVM → future capabilities. Besu's EVM makes asset, payment and custody functions directly possible with mature ERC patterns (in Cosmos this is an add-on layer).
- Same stack as EBSI. EBSI uses Besu + IBFT 2.0 (RS-EBSI-0001). A common base eases interoperability and the transfer of know-how.
- Single-chain simplicity. Cosmos's multi-chain/IBC advantage would sit idle under Tamga's "one shared chain" decision.
- Built-in fit. Permissioned operation, PoA/equal votes and immediate finality come out of the box with Besu QBFT.
- Full sovereignty. Apache 2.0, our own network, forkable; no dependency on any company or network.
Consequences
Positive:
- A ready foundation for the asset/payment roadmap (EVM).
- Technical proximity to EBSI → easier interoperability.
- A mature tooling and developer ecosystem.
- Simple, single-network operations.
Caveats / cost:
- QBFT struggles above ~20 validators → addressed by "states = validators, institutions = full nodes"; if the number of states grows very large, a scaling strategy is needed (
PM-BC-0001, Future). - Requires Java-based operations expertise.
- Lower TPS than Cosmos, but sufficient for a trust-infrastructure workload.
Alternatives considered
- Cosmos SDK (CometBFT): better validator scaling and TPS; but no native EVM, the multi-chain advantage is idle for Tamga, and a different stack from EBSI. Rejected.
- Consensus from scratch: already rejected in
PM-BC-0001(risk, time, no contribution to sovereignty). - Building on Ethereum L1/L2: rejected in
PM-BC-0001(loss of sovereignty, GDPR exposure).
Related documents
PM-BC-0001— this ADR closes the open "layer 4" decision there.RS-FRAMEWORKS-0001— the comparison this decision rests on.- RS-EBSI-0001 — the EBSI precedent (Besu + IBFT 2.0).
- ACA-BC-0001 — consensus fundamentals.