delvingbitcoin

Batched Splicing Considered Risky

Batched Splicing Considered Risky

Original Postby ajtowns

Posted on: November 8, 2023 19:35 UTC

The discussion revolves around the complexities and potential vulnerabilities associated with channel management in blockchain networks, focusing on scenarios involving double-spending, splicing, and channel reopening strategies.

One critical point raised is the possibility of a party, referred to as B, exploiting the system by using a revoked commitment transaction to double-spend within a pre-spliced channel. This maneuver raises questions about the necessity and security of choosing between a normal unilateral close, utilizing the current non-splice commitment transaction, and the mechanisms to regain access to funds for further transactions with other parties, such as C and D.

The conversation further explores the implications of a naive action taken by D, who proceeds to send the preimage to the Hash Time-Locked Contract (HTLC) for the Just-In-Time (JIT) channel payment upon witnessing the broadcast of a batched transaction. This scenario underlines the risk associated with forwarding HTLCs without considering the suitability of the channel for zero-confirmation (zeroconf) use, especially when not all inputs are controlled by the forwarding party. The dialogue suggests exploring advanced versions of zeroconf channels that could accommodate entirely new funding transactions without necessitating the closure and reopening of the channel, though it acknowledges the current limitations in achieving such functionality.

There's also a discussion on the need for flexibility in managing channel funds and relationships, especially in situations where a splice operation involves changing participants or funding sources. Specifically, the ability to revert to original channel funding outputs without closing existing channels is highlighted as a desirable feature. This becomes particularly relevant when a splice transaction is rendered obsolete due to the prior spending of a targeted unspent transaction output (UTXO), indicating that dropping the splice transaction under these circumstances should be an acceptable resolution.

In summary, the conversation touches upon several intricate aspects of channel operations within blockchain frameworks, including the challenges of preventing double-spending in channel splicing, the technical considerations for safely forwarding HTLCs, the potential for advanced channel management features to support dynamic funding arrangements, and the procedural nuances involved in adjusting splice transactions following UTXO spending events.