Non-confiscatory Transaction Weight Limit

The discussion presents a nuanced debate on the handling of large transactions within a blockchain framework, specifically addressing concerns related to transaction size and block stuffing. A significant point of contention is the proposal to treat every transaction as if it had approximately 1 MvB in size, with particular scrutiny on the treatment of transactions greater than 100kvB. The concern is that this policy might incentivize users to fill their transactions with unnecessary data, as they would be charged for 880kvB extra space regardless of the actual size of their transaction. This could potentially lead to inefficiencies and bloated blocks.

A novel approach is suggested to mitigate these concerns while still managing the size of transactions efficiently. The proposal includes specific consensus rules: allowing both the coinbase transaction and the last transaction in a block to be of arbitrary size, while imposing a weight limit of less than 400,000 for all other transactions. If the last transaction exceeds this weight limit, it faces certain constraints such as not being able to spend an output created in the same block and having its weight rounded up to facilitate calculation and avoid partial block filling. These measures aim to optimize the use of block space by adjusting how transactions are weighted and included in blocks.

For mempool acceptance and storage, the proposal introduces a system where large transactions are categorized into 36 buckets based on their adjusted weight, simplifying the selection process for miners by only considering the highest fee transaction in each bucket. This system ensures that only the most economically viable transactions are kept in consideration for block inclusion, thereby optimizing the mempool management.

The mining strategy outlined involves selecting large transactions from these buckets in descending order of size and attempting to include them in a block without exceeding the block's maximum weight limit. The algorithm provided aims to maximize fee collection while respecting the block's weight constraints, even though it acknowledges potential limitations in achieving the optimal block composition.

This discussion highlights the complexities involved in balancing transaction efficiency with network health, proposing a detailed mechanism to address some of the inherent challenges in managing large transactions within a blockchain environment.