Scaling Lightning Safely With Feerate-Dependent Timelocks

Firstly, he challenges the rationale behind the alignment of 'window' in the context of blockchain mining. He points out that when a 'window' is aligned—meaning the consideration of a set number of blocks that ends with the most recently mined block whose height is divisible by the window size—it impacts how current fee changes are reflected in transaction validity. In essence, a recently increased fee may not influence whether a transaction is confirmed because only the blocks within the aligned 'window' are considered for assessing validity.

Subsequently, Boris proposes an intriguing alternative: the implementation of a global window_size. This concept would streamline the process by eliminating the need to account for non-aligned windows, thus simplifying verification processes. Moreover, it would lead to space efficiency within the Full Data Transaction (FDT), as only a single array of integers representing the window size would be necessary. This array would be capable of answering queries regarding the presence of a minimum number of blocks within the latest blocks that have a median fee rate equal to or greater than a specified threshold.

Boris's questions highlight potential areas for optimization in blockchain technology, specifically in how transaction fees are assessed and how data storage can be minimized without sacrificing the integrity of transaction verification. Such discussions are integral for the ongoing development and refinement of blockchain systems.