Highly Available Lightning Channels Revisited – ROUTE OR OUT

The discussion revolves around the critical importance of achieving perfect reliability in payment success rates for the Lightning Network to be considered a successful payments platform. Despite this, it's argued that pathfinding, as it stands today, is not a significant barrier to achieving this goal. The substantial efforts already invested into improving pathfinding, including regular probing and the use of sophisticated algorithms detailed on Lightning Dev Kit's blog, have resulted in payments generally succeeding on the first attempt, assuming the payer is willing to cover the necessary fees. This success is attributed not only to the robustness of the current pathfinding logic but also to the strong connectivity of the network, which is expected to scale to larger networks with the aid of background probing and potential future enhancements like trampoline routing.

However, payment failures do occur, predominantly not because of pathfinding issues but due to other factors. Such failures often result from nodes not conducting regular probes or failing to fetch scoring information from reliable sources, actions deemed essential for all nodes. Insufficient liquidity is another common cause of failure, whether on the part of the recipient, who may lack Just-In-Time (JIT) channel services, or the sender, who might not possess adequate capacity due to various technical and operational limitations. These challenges are being incrementally addressed, notably with the development of new channel types like TRUC, which aim to mitigate these constraints.

The proposal under scrutiny is critiqued for potentially introducing significant social costs to the network. Specifically, it risks incentivizing senders to bypass improving their pathfinding solutions in favor of routing payments through High Availability (HA) Nodes. This binary choice places average node operators in a dilemma: either they do not signal as "HA" and miss out on routing volume, signal as "HA" but risk losing credibility and volume if they occasionally cannot meet liquidity demands, or engage in JIT rebalancing at the cost of higher fees, which may still not guarantee payment success and could lead to a loss of routing volume. Moreover, such a system could disproportionately benefit large custodial operators capable of facilitating immediate rebalancing through zero-confirmation channels among themselves, thereby centralizing control and undermining the decentralized ethos of the Lightning Network.

In conclusion, the argument posits that before adopting measures with vast social implications for the network, there should be a comprehensive evaluation to ascertain their necessity and explore alternative solutions to the underlying issues. The current state of affairs, where non-fundamental issues are conflated with those critical to the operation of lightning routing, suggests that the proposed changes are premature and potentially misguided.