Published on: 2022-12-13T12:55:08+00:00
The bitcoin-dev mailing list is currently engaged in a discussion about Full RBF (Replace-by-Fee) and its potential impact on mempool policy. Advocates for Full RBF argue that relying on predictable mempool policy is not advisable, as miners will always prioritize optimizing fees in the short term. This could lead to a scenario where miners advertise how to connect to their nodes, allowing those who prefer higher fee transactions to send them directly. However, this assumption is not an economic fact of life, as there are alternative options such as stablecoins or ETH that users could switch to, resulting in a substantial drop in bitcoin traffic, transaction fees, and price.The outcome of implementing Full RBF is uncertain in terms of its effect on the desirability of bitcoin and its price. The use of BTC for payments may shift towards lightning network, causing a reduction in on-chain traffic and fee income. The author emphasizes the importance of prioritizing long-term income over immediate revenue to prevent 51% attacks from becoming a dominant strategy. Consistent and predictable relay policy across nodes remains possible regardless of whether the policy follows the "first seen is final, highest fee rate wins" principle or something else. The author sees Full RBF as a way to eliminate a use-case that carries risk, ultimately strengthening Bitcoin. However, they disagree with preventing others from voluntarily taking on small risks that can be easily managed or avoided. The author clarifies that "money for enemies" refers to empowering the majority rather than an elite minority within the Bitcoin community.Regarding the claim that RBF is a fee-minimization feature, the author responds by highlighting the economics of the situation. Fees serve as a price on confirmation, and there exists an optimum price where total earnings vs price reaches a peak. RBF optimizes the discovery of this optimum price, which is desirable. However, many merchants and consumers reject opt-in-RBF, necessitating the need for full-RBF to improve price discovery of blockspace when such acceptors are prevalent. The author also disputes the implication that demand follows supply, asserting that if Bitcoin offers less utility, it will experience reduced demand regardless of its optimization for capacity saturation. The optimal state is achieved by providing the most value per blockspace per time period. Zero-conf transactions currently generate added demand for blockspace from merchants and consumers due to their qualification for zero-conf acceptance based on fee rates.A discussion on full RBF highlights its alignment with miner and user economic incentives. However, this argument can be refuted as RBF primarily serves as a fee-minimization feature, resulting in lower earnings for miners. Additionally, nodes providing replacement service and propagation incur expenses. The author emphasizes that fees are the price of confirmation and applies the concept of "price theory" where an optimum price may be lower than the prevailing market price. Therefore, RBF optimizes the discovery of this optimum price, but many zero-conf acceptors reject opt-in-RBF despite its improved price discovery. Thus, full-RBF becomes necessary in such cases.There exists a fundamental disagreement regarding Full RBF. Supporters perceive it as a means to eliminate the risk associated with zero-conf transactions, while opponents believe that relying on predictable mempool policy has worked well thus far. Those in favor argue that node policy is not a consensus rule, and changing mempool policy is essential to actively discourage zero-conf. Opponents contend that zero-conf acceptance is already widely used and valuable, and full-RBF can be discarded if the majority of node operators do not care about it. The economic incentive argument currently holds little significance, and there remains an obvious incentive for someone to attempt a double-spend attack on a zero-conf merchant. Angus suggests that accepting zero-conf changes from being tolerated with occasional issues to being strongly discouraged due to potential financial losses. For now, Angus supports running a full-RBF node, considering it a measure that strengthens Bitcoin. However, he is open to reversing his stance if a strong economic argument convinces him and other miners or users to oppose full-RBF.While many proponents of zero-conf focus on supporting Lightning as an alternative, the author argues that Lightning is not without risks and should not be solely relied upon for commercial payments. Lightning is suitable for low-value payments due to the current low-fee environment's lack of scalability. However, for high-value payments, zero-conf has never been valuable or useful and likely never will be. The author highlights instances where double-spends of zero-conf transactions without RBF have been repeatedly demonstrated, emphasizing the absence of a long-term scenario where zero-conf holds value. Although low fees may seem appealing, they incentivize network-wide surveillance, increase the burden on nodes, and promote unsustainable practices resembling a "mev" (miner-extractable value) bounty for merchants. Nevertheless, the author identifies the risk involved in waiting 20 minutes before shipping items as insignificant, with the only potential issue being a user replacing a transaction with a different destination.
Updated on: 2023-08-02T08:35:26.336429+00:00