Bitcoin Optech Newsletter #432
Bitcoin Optech’s #432 newsletter examines decentralized mining‑pool payout designs and a new denial‑of‑service flaw in Eclair. It finds that current payout schemes face scalability attacks and that a simple protocol tweak can boost pool revenue by ~0.2 BTC per year.
Key points
- Decentralized coinbase payout protocols (DCPPs) struggle with firmware limits and rich‑list attacks.
- A “value‑weighted random draw” or operator‑reserved outputs are the only viable alternatives today.
- Eclair 0.14.0 contains a memory‑exhaustion DoS bug fixed in 0.14.1; users should upgrade immediately.
- Stale‑tip relay of full block data could raise a 20% pool’s annual earnings from 16.1 BTC to 16.36 BTC.
- P2Pool’s growth alone pushed out miners earning < 0.001 BTC per block, prompting a 16‑output limit proposal.
- P2Pool tx examined
- 2,545
- Small miner threshold
- 0.001 BTC
- Proposed output limit
- 16
- Pool hashpower share
- 20%
- Current stale‑relay earnings
- 16.1 BTC
- Optimized stale‑relay earnings
- 16.36 BTC
Decentralized payout challenges
Bitcoin Optech highlights a recent analysis by developer 7d5x9 on how decentralized mining pools might pay miners directly from the coinbase transaction, a concept dubbed decentralized coinbase payout protocols (DCPPs). The core problem is that each miner‑specific output inflates the coinbase size, which is capped by mining firmware and displaces fee‑paying transactions. When a pool has more miners than outputs, any work‑ranking rule enables a "rich‑list attack" – a large miner can split its hashpower across many identities, outranking smaller participants and forcing them out of the payout.
Proposed mitigations and limits
The report notes that earlier designs merely shift the attack surface. Braidpool, for example, moves control from output selection to the pool’s threshold key. Without a soft‑fork enabling rebindable signatures (e.g., BIP‑448), a single key or threshold retains control over the sole coinbase output. 7d5x9 sketches using Spark‑style statechains or Ark, but these still allow an attacker to block cheap cooperative settlement by holding a small claim per subtree. Consequently, the analysis concludes that a join‑pool cannot decide which outputs appear in the coinbase; the only practical options are a value‑weighted random draw (attributed to Matt Corallo) or reserving outputs for designated operators. LearnBitcoin.com’s review of 2,545 P2Pool coinbase transactions supports this, showing that growth alone squeezed out miners earning < 0.001 BTC per block. For P2Pool v2, the author jungly proposes capping outputs at 16 given current firmware constraints.
Eclair DoS vulnerability
Another focus of the newsletter is a memory‑exhaustion denial‑of‑service bug in Eclair 0.14.0, disclosed by rorp. The flaw stems from a missing lower‑bound check on the BOLT‑2 "start_batch" message’s batch_size field. An attacker could send batch_size = 0, causing the node’s internal list to grow without limit and eventually crash the process. While the bug does not expose funds directly, repeated crashes could jeopardize channel balances. The issue was patched in July 2026 with release 0.14.1, and Bitcoin Optech urges all operators still on 0.14.0 to upgrade.
Stale‑tip relay debate
The newsletter also recaps ongoing discussion around BIP‑332 stale‑tip relay. Antoine Poinsot argues that relaying only block headers suffices for making stale blocks observable, whereas full block data speeds up reorgs and can increase a miner’s revenue. Anthony Towns estimates that a pool controlling 20% of network hash rate currently earns about 16.1 BTC per year by mining on its own stale block, and that reducing the relay delay from five seconds to one could raise earnings to 16.36 BTC. While the incremental gain is modest, the debate underscores trade‑offs between network efficiency and miner incentives.
Implications for the ecosystem
Overall, Bitcoin Optech’s findings suggest that truly decentralized payout mechanisms remain technically constrained and vulnerable to strategic attacks. Until protocol upgrades like BIP‑448 land, pools may need to adopt hybrid schemes—random draws or operator‑reserved outputs—to preserve fairness. The Eclair fix serves as a reminder that even well‑maintained implementations can harbor critical bugs, reinforcing the need for rapid patch adoption. Finally, the stale‑tip relay conversation highlights how subtle protocol tweaks can affect miner economics, a factor worth monitoring as the network evolves.