Umberto | Mostly Data 🧙‍♂️

@brian_smith_0 @minnus I think that first one should remove the bias. I'm looking into this since November 2025...and behaviour is pretty much the same since then. By removing location of vote txs the difference is much more nuanced dune.com/queries/622544…

@minnus Will be interesting to rerun this in a week with maker priority changes captured

@StakingMatthias lol

Highest IBRL score among the top 10 highest-staked Solana validators: Staking Facilities Built through long-term focus on reliability and steady improvements If you care about Solana’s future, it matters who you stake with ibrl.wtf

@minnus @0xTaker are you reindexing by removing vote txs?

A better oracle update placement chart, highlighting the lack of uniformity across Solana block builders. For each propAMM update, we measure where it lands in the block as: tx index / total txs in block Then bucket the result into 5% slices.

@vitorpy @cavemanloverboy Innovation?

@MostlyData_ @cavemanloverboy it's facts and the italians can't cope with american innovation

the italians were right about a lot of things

@vitorpy @cavemanloverboy I’m going to press it

@cavemanloverboy except for pineapple pizza. pineapple pizza is legit and that's a hill i'm willing to die on

@toly What is Ethereum?

We have pqc before ethereum. Quic, FUD Ethereum

What people mostly miss in the FBO vs FBA vs FCFS debate is that in a distributed system you have an additional player: the validator. This change the equilibrium. FBO does not eliminate the rent from stale state. It turns that rent into an auction object: the right to move the state first. This distinction matters. In FBO, the price is not updated through a single batch-level clearing step. The state price updates transaction by transaction, according to the ordering chosen inside the batch. Paying more priority lets a participant buy an earlier position in that state transition sequence. So FBO can be better than FCFS if it replaces an opaque latency race with an explicit priority market, and if the resulting rent can be redistributed to the protocol. But FBO can also be equivalent to, or worse than, FCFS if the validator or builder can appropriate the auction, accept side deals, censor competing bids, or self-preference their own transaction. It can also reintroduce a latency race if priority bids are observable and continuously updatable until the batch closes. In that case, the race is no longer only about who arrives first, but about who can update their bid last. So you can see why the combo MCP+FBO is a powerful combo if designed properly.

@wintermute_t Hack

In 2020, we had DeFi Summer In 2026, it’s _____ Summer

@raikucom @Cetipo I think @Cetipo just misunderstood what was happening and this opened a wave of conjectures. x.com/buffalu__/stat…

My assumption is that by refusing to include any other transactions, they're trying to ensure minimal state changes to maximize the odds of being able to include any and all high-value txs/bundles even if they were to arrive during the last tick of the block Refusing to update state also creates additional price dislocations between CEXs and onchain markets, in a way generating their own artificial demand for high-value transactions

solscan.io/block/416414661 Only 8 tx in this block. Validator trying to maximize its profit by only landing tx with fat jito tip.

in case you're curious how rare it is to have a block with >0 AND <10 non vote transactions, on median it happens 65 times/epoch or 0.015% of the time

@jarxiao @solana just seeing this. issue was on our side in BAM that we've been running down. these were the highest reward/cu txs at the top of the block, so they naturally got selected first.

@lbkolev @Cetipo probably I have to revisit my math knowledge, but $ deal + anything else > $ deal

@MostlyData_ @Cetipo unless u have an under the table deal for more $ than the cumulative sum of all (potential) tips from the other txs

@Austin_Federa I think this is the first time in my life I've seen someone stop at a stop 😅

Hired the world’s best driver today

@toly @MaxResnick 100% Mine was just a small observation on that chart…I would remove that 😅

@MostlyData_ @MaxResnick If we migrate to pqc, that gives us the right to FUD everyone who hasn’t yet, and that’s the real benefit of pqc.

If physical error rates continue to decline on trend for rubidium based neutral atoms architecture, a quantum computer capable of breaking Ed25519 signatures could be ready is as little as 5 years on the most aggressive timeline. Q day is most likely still 10-15 years away but there is a chance it could come sooner. Read more about how we plan to secure the Solana protocol against a powerful quantum adversary in our new blog w/ @samkimcrypto. Work has already begun on some of the preliminary aspects of transition like increasing transaction sizes and adding a FALCON verification system call. This blog post goes into the other aspects of the protocol that need to be upgraded as well as migration strategies.

@IgMosqueira @dktrpo @toly Can you argue on why FCFS is a superior product? And what makes FBO bad? Plus, MCP is not a market microstructure feature, you are making some confusion here

Yes... Plus I am not saying that there couldn't be a sensible crypto design. But MCP doesn't make you better than Nasdaq no matter what Toly says But whatever design you come up with has to survive the fact that arbitrageurs will always pit clobs against each other so saying Frequent Batch Auctions compete in price whereas continuous time trading competes in latency sounds good but it is wrong as currently proposed...

As is his custom @toly obscures what is already misleading. Two simple points for the braniacs out there: 1) if Budish is modeling this correctly, then what is the optimal batch auction delay in his view 2) if batch auctions remove the need for latency... x.com/toly/status/20…

If I have to quantify, 80% yes 20% no. There are some nice properties of FBO that FCFS doesn't have. Solana largely proved that you can have low spread and high liquidity on spot even w/o continuous markets. So I don't buy the story that FCFS is superior because in tradFi it works better.

@MostlyData_ @IgMosqueira @toly But you do agree that FCFS would likely be superior, if we _could_ have a fair implementation of FCFS in Byzantine settings like blockchains?

@alessandrod Right, you don’t know how bad it can go…

@MostlyData_ 🤣🤣🤣 hey at least it isn’t a cappuccino

@alessandrod Like drinking a spritz while eating a caprese…

before the americans come and ruin summer