letsgetonchain
@letsgetonchain
Followers
2K
Following
6K
Media
207
Statuses
1K
Breaking down the standards and mechanisms shaping crypto, mostly DeFi, sometimes infra.
Joined September 2024
🚨ETHEREUM UX IS ABOUT TO 100X YOU ARE NOT READY 🚨 EIP-7702 is about to be shipped! I explain how it works and go through every line of code in the evm (revm) that changes because of it 🤓 No better way to really understand it!
I'm so excited about EIP 7702 that I am going to spam your timeline with a video about it tomorrow I will show you every new line of code in the evm that comes with it 🫡
42
85
512
Maybe this is obvious to some, but digging into @Lighter_xyz finally made me understand how zk rollups actually bind the data they post on Ethereum's DA layer (via blobs) to the state transitions they prove.. Here is how this works and why its important 🧵
5
16
109
happy to share my ref link for lighter. Lighter is on the right path to power truly trustless perps build on Ethereum! https://t.co/mFNybviwMM
app.lighter.xyz
Lighter - Trade perpetuals with unmatched efficiency and security
0
2
3
7/ Result: The proven state transition is cryptographically tied to the posted blob. Hence for every verified state transition you can reconstruct the account and market state of Lighter to compute balances, pnl, funding, pool shares, and ultimately generate proofs to force
1
0
7
6/ to recap: 1) Posted blob == claimed blob ✅ 2) Execution data == claimed blob ✅ --> posted blob == execution data ✅ (which was the goal)
1
0
8
5/ Part 2 -> Execution Public Data == Claimed Blob The public execution data here refers to everything the provers circuits touch that has to be publicly reconstructable which should be equal to posted blob data (like mentioned above account delta + market data). Inside the
1
0
8
4/ Lighter ensures that the blob data it posts is exactly the data from execution through a two-part binding mechanism: Part 1 - Posted Blob == Claimed Blob "Claimed blob" here refers to the blob bytes the prover says they used inside the proof. A blob's data can be
1
0
9
3/ Why does that matter? If the sequencer stops operating, you must be able to reconstruct the state from this blob data to generate a proof to withdraw your funds on Ethereum. That is what makes the system self-custodial instead of "trust me bro, your funds are safu".
2
0
12
2/ First off, what is does Lighter post as blobs? Lighters blobs contain: - all account changes, - market data such as index/mark prices used to calculate funding, liquidations, pnl etc. If you download every blob you can rebuild the account/market state of the exchange.
1
0
7
1/ The question: how does Lighter prove that the state commitment is not only accurate but also that it can be reconstructed from the data it posts to Ethereum as blobs?
1
0
6
Maybe this is obvious to some, but digging into @Lighter_xyz finally made me understand how zk rollups actually bind the data they post on Ethereum's DA layer (via blobs) to the state transitions they prove.. Here is how this works and why its important 🧵
5
16
109
1/ The question: how does Lighter prove that the state commitment is not only accurate but also that it can be reconstructed from the data it posts to Ethereum as blobs?
0
0
1
The reason a perp dex like @Lighter_xyz is the perfect fit for a an Ethereum rollup (despite current DA constraints) is that it does not need much DA. Only state diffs are needed to rebuilt the state and exit the system. >99% of orders never match and don't consume any DA.
This is actually a question I've had without digging into the data. → Lighter is a rollup, which means it should be using Ethereum blobs. → Blobs have shit capacity and were basically maxed out pre-lighter. → Base is doing 3% of Lighter's TPS but generating 2,500% the
0
1
14
This is no solvency issue but a liquidity one. Hopefully the learning of this is that we get better at pricing liquidity risk premiums for these isolated markets. Advantages of pooled models like Fluid or Aave are that borrowers dont have to pay that risk premium to the
@sunboud0 @MorphoLabs @hyperithm ok let me be specific @MEVCapital USDC = $225M stuck Smokehouse USDC = $199M stuck Hyperithm USDC = $97M stuck Gauntlet Frontier = $63M stuck Alpha USDC = $45M stuck Clearstar = $43M stuck Gauntlet Core = $32M stuck Relend = $13M stuck Steakhouse infiniFi = $10M stuck Edge
2
2
26
I see platforms like euler or morpho as infra providers who use their own infra. But they should not have their brand's risk management be associated to other risk management practices. thoughts @euler_mab?
1
0
11
Modular credit markets are not the problem. The problem is blurred responsibility. @eulerfinance and @MorphoLabs should only display their own curated markets within their branded UI. Let external curators distribute via their own UI and have third party UIs aggregate all
This is the reason I've been skeptical on the curator based model for DeFi lending & borrowing protocol. The platform are taking on the reputation risk and duty to care for their users whether they like it or not and a few curators that went rogue with poor risk management will
17
5
121
The first edition was my entry into crypto Def ordering 2nd edition in paper
Mastering Ethereum 2nd edition is officially out. How can you read it? - Online for free: in the next few days/weeks we will publish it on github/x and probably other venues - Kindle: you can buy it on amazon - Paperback: you can buy it on amazon - Online not for free (doesn't
1
0
7
Im diving into a two month research project with @cyberfund on perps I want to include examples for various designs to compare/contrast trust assumptions etc. Architectures I am planning to include: CLOBs @HyperliquidX as example of native L1 DEX @Lighter_xyz as example of
7
1
18
Reading @l2beat summary of @Lighter_xyz. Is there any perp DEX that allows for trustless order timestamp assignment today? L1s like @HyperliquidX give validators this power. Other perp DEXes give this power to the operator/sequencer. What are the most promising solutions
0
0
2