Batch exchange withdrawal to lightning requires covenants

Antoine is responding to Bastien's comments and acknowledging that he now understands the concept of regular splice, where one participant can double-spend at any time with a previous commit transaction. He mentions that this liquidity griefing issue is already known and advises waiting for a few confirmations before using spawned channels.

Antoine then discusses the scenario where the splice is a "standard" nVersion=2 transaction with a reasonable feerate for non-delayed broadcast. In this case, his previous comment does not apply and he agrees with Bastien.

However, Antoine raises a concern about a hypothetical future situation where the feerate picked up for the batch splicing is not compelling enough during mempool spikes. He questions whether it would be possible to interactively re-generate a bumped Replace-By-Fee (RBF) transaction in such a scenario. He expresses concerns about the impact on rebroadcasting the batch splice transaction package with a bumped Child Pays for Parent (CPFP) fee. Antoine states that this concern needs to be verified.

Overall, Antoine's email acknowledges Bastien's comments and provides further insights into the topic of regular splice and its potential implications in different scenarios.