Scaling Lightning Safely With Feerate-Dependent Timelocks

The discussion led by Eric Voskuil on the bitcoin dev mailing list raises significant concerns regarding the future of Bitcoin's decentralization and its resistance to censorship. The efficiency of paying additional miners out of band is expected not to scale favorably compared to the likelihood that a user’s payment will be processed in a timely manner. Voskuil points out the diminishing returns of engaging with smaller mining operations, as securing transactions within six blocks can be virtually guaranteed by reaching out to miners representing 90% of the total hash rate. This assurance diminishes the incentive to connect with the remaining 10%.

A critical issue mentioned is the potential shift in mining revenue towards out-of-band payments, which could disproportionately benefit larger miners and incentivize further centralization of mining operations. Such centralization poses a threat to Bitcoin's core value proposition of censorship resistance. Although preventing out-of-band payments outright is not feasible, Voskuil stresses the importance of maintaining in-band fee payments and transaction relays as the most effective methods for users to secure their transactions. Ensuring these remain the preferred route is crucial to preserving Bitcoin’s resistance to censorship.

The reliance of Layer 2 solutions like the Lightning Network (LN), coinpools, and channel factories on the uncensored nature of Bitcoin transactions is also highlighted. These technologies presume that transactions will not face censorship, implying no additional security risks when designing them to depend on accurate in-band fee statistics. Finally, Voskuil suggests that issues such as miners prioritizing their own transactions or accepting occasional out-of-band fees can be mitigated through user strategies. These include employing fee-dependent timelocks with extended time windows and setting feerates below the maximum threshold they are willing to pay, thus deterring any negative implications from varying local transaction selection policies by miners.