Author: Anthony Towns 2023-07-26 08:05:23+00:00
Published on: 2023-07-26T08:05:23+00:00
In an email thread, participants discussed various aspects of Bitcoin network policies and potential solutions to address certain issues. One topic of discussion was the "top of mempool" assumption. It was noted that there are multiple goals to consider: transactors want their transactions to be included in blocks, pools/miners want to see profitable transactions quickly, node operators want to support users while avoiding unnecessary resource usage, and those concerned about decentralization want fair transaction selection without monopolistic advantages.Regarding lightning unilateral closes, it was suggested that special cases could be made for them to improve the achievement of these goals. Lightning nodes could relay these transactions even when regular nodes don't, and some miners could use specific configurations to include these transactions.The question was raised whether there could be a future where miners don't care about policy restrictions at all. It was argued that, from a mining perspective, policy restrictions are not necessary and mainly serve other purposes such as enabling efficient node operation and promoting decentralization.The concept of zero-fee transactions with ephemeral outputs was discussed. It was explained that making such transactions zero fee prevents them from being mined in non-attack scenarios, which avoids creating dust unspent transaction outputs (UTXOs). However, it was pointed out that an attacking miner can already create arbitrary dust UTXOs regardless of fees.The addition of trimmed HTLCs (Hashed Time-Locked Contracts) to the ephemeral anchor was considered. Concerns were raised about keeping values in OP_TRUE outputs being taken. It was clarified that value in an OP_TRUE output can only be taken by confirming the parent transaction, just like if the value had been spent as fees. Directly putting the value into fees would violate the constraint that ephemeral output transactions must be zero fee.A hybrid approach to channel jamming mitigation was proposed. The approach involved implementing monetary measures in combination. The idea was to make the cost for attackers scale with their usage, while regular users pay reasonable fees. This could involve charging for bandwidth and compute resources used in message spam, imposing fees for holding funds in liquidity DoS prevention scenarios, and setting higher fees for actual payments to incentivize their prioritization. The profitability of routing traffic and the ability to spin up new nodes at lower fees were mentioned as factors that can help balance the system.Concerns were expressed about the determination of small fees and the potential for market influence. It was argued that if fees on the network are high, it would be easy for new lightning nodes with slightly lower fees to attract traffic. Thus, the market dynamics would help regulate fee levels.The challenge of proving time in case of issues and the reliance on cooperation were discussed. It was noted that since the amounts involved are relatively low, precision beyond an hourly basis may not be necessary. Moreover, a universal clock is not required, and per-block measurements and reorg/stale block considerations could be sufficient.A scenario involving a specific routing path and payment expiry was analyzed to highlight potential issues with fees and channel closures. It was concluded that this scenario is unlikely and can be addressed by imposing caps on liquidity fees based on the route position and expiry time.Finally, the participants confirmed their resolution on the need for a hybrid approach to address channel jamming. No further questions or alternative solutions were raised.Overall, the email thread provided insightful discussions on various aspects of Bitcoin network policies and proposed solutions to mitigate channel jamming and improve transaction selection and efficiency.
Updated on: 2023-07-27T02:07:39.327518+00:00