A fairer trading plan of Solana blocks because the block producers can still choose their incoming orders.

What is this proposal?

A report from CryptoSlate states that the codebase request for SIMD-0649 was closed on September 25 without having been incorporated into an actual rule. The original proposal required a block checker or validator to reject blocks if transactions within the same batch of lists or entries were recorded by unordered high-to-low or equal priority order, but did not specify which transactions should be in those blocks, leaving only issues for further discussion rather than enforcing immediate reordering rules.

Why is it so important to order?

Order senders on Solana want to predict which order their orders will go in, because that order affects price and comparative advantage. The body therefore focuses on verifying whether the non-exempted transactions selected by a blog creator or leader within the same group are listed according to specified scores.

The rules don't line up all the slots.

The structure dictates that the creator of a blog or leader can also pick and choose transactions, group listings, move deals to later groups, and even select end-of-group settings. Intellectual verification covers only order within an entire group so it's checkable and enforceable, not sorting through slots, not ensuring that the most important orders always come first, and not guaranteeing best execution.

How is the priority score calculated?

The sorting account of Solana combines the listings into groups. In this form, businesses that are not excluded in each group must appear in non-increasing order of importance scores; an auditor reviewing a block will compare recorded scores and assume errors to be incorrect blocks but will not automatically switch orders after receiving a block.

A priority score is calculated from the returns that a leader receives for incorporating a business, divided by the requested cost in accordance with the pre-operational capital model. The body defines its calculation as an integer, multiplier and one unit insurer in terms of factors so that the computing agent matches each other. According to the customary rules described by the body, the return combines the priority fee and the unspent part of the base fee, thus being more specific than ranking based on user-specified fees alone.

The gap the body's trying to close.

The leader is prevented from making the rule meaningless by creating list groups that are so small there's nothing to compare them with, thus dictating that a pre-final group must contain at least two FEC packets called shreds according to the FEC size rules of the reference frame. This means that there should be at least 64 data shreds and the final group still needs to pass through sequence checks but has an exception in terms of minimum sizes because slots may end before they fill up.

The argument is not over.

In a September 23 review, one reviewer argued that leaders may still close groups when it is in their interest to separate conflicting businesses and ask for current group size information separated by queue organizer, client and market conditions as well as testing speed versus minimum sizes. This points out the limitations of the rule in practice but not evidence that leaders use this tactic on mainstream networks.

It also notes that Agave and Firedancer aim for groups of about 2 FEC sizes, but there is no actual measurement on how often the current leader creates smaller groups, so it's not possible to estimate the effect of minimum size on normal block production.

August's discussion also worried that waiting for the whole group to check could interfere with Firedancer's way of partially re-checking incoming data. The revised body thus allows a checker to compare and transact when the data arrives, then reject the block if subsequent comparison fails; but it also acknowledges that allowing the leader to wait until two FEC minima have been completed may increase time to distribute data in low reporting network conditions where the source does not measure latency.

The SIMD-0649 does not prevent the leader from benefiting their own business by paying a premium to themselves, because those fees go back to the leader while the burned portion of the base fee is still capital.

The shutdown of the codebase on September 25th came after a call for debate and support from additional client developers.

In terms of cost and speed.

Related reports also cited the Solana issue reversing Ethereum in fees, while ETH was still leading on burning issues including a 300ms speed debate to compete with trading bots and concerns that speeding up 250ms could trigger network instability.

What Should Readers Know?

The leader's choice determines whether two competing businesses will meet the same sequence of tests, for example, a business with high scores in the latter group will not be promoted ahead of a lower-scoring business in the former group simply because it has higher scores.

The point that can change is the clarity of internal hierarchy control, not a guarantee that businesses who should have been there in the first place will always be there. If an amendment to the rule were progressive, value for users would depend on how often meaningful competitors stayed within the same group and whether or not minimum size changed leader behavior. Public proposals specify how to check relative ranking data but lack of group size leaves uncertainty about actual performance.