Stop looking for a partner. Focus on your goals. Focus on loving yourself. Focus on growing as a person. Rebuilding your life. Then the right person will find you. - Unknown.
"My friends think I'm a crypto genius, (I'm not) it's because I read Coin Club’s weekly newsletter (and they have no idea it exists)" - Every Crypto Degen
This one's for the Crypto Whales ⬇️
NEW exclusive & luxurious experiences have been added to https://t.co/6tQs9fsGkP. Apply today & #TravelUnlikeAnyOther with your favorite cryptocurrencyThis one's for the Crypto Whales ⬇️
NEW exclusive & luxurious experiences have been added to https://t.co/6tQs9fsGkP. Apply today & #TravelUnlikeAnyOther with your favorite cryptocurrencyThis one's for the Crypto Whales ⬇️
NEW exclusive & luxurious experiences have been added to https://t.co/6tQs9fsGkP. Apply today & #TravelUnlikeAnyOther with your favorite cryptocurrencyThis one's for the Crypto Whales ⬇️
NEW exclusive & luxurious experiences have been added to https://t.co/6tQs9fsGkP. Apply today & #TravelUnlikeAnyOther with your favorite cryptocurrency.
$INJ just smoked target 1 - Trimmed half the trade and now we are free riding!
Join us... stay for 6 months and get our 360 Elite Traders Course for FREE ($2499 value)
Metis is not only the #Ethereum Layer 2 with the lowest fees, Metis is also an ecosystem full of fascinating dApps, with DeFi as its bread and butter🤌
Check it out (https://t.co/K8oOrfOnmT) and try it out (https://t.co/U7ra491m0K)😎
BREAKING🚨
St. Galler Kantonalbank, a major Swiss bank, just announced they're stepping into the crypto space with #Bitcoin and #Ethereum custody services!
Here are the latest updates on bringing statelessness to Ethereum, fresh from yesterday’s Verkle Implementers Call 🐘🔥
TL;DR: we have a new testnet (Kaustinen) live, with 4 clients and counting. Work is underway on integrating Verkle into Geth, which will happen over the next few months. Several other client teams are making significant progress as well. Still early, but it looks promising that the performance of Verkle will be competitive with the current Merkle Patricia Tree. There is a shadowfork in the pipeline as well, which will help verify our current assumptions.
Read on for the full update below.. And hop into #verkle-trie-migration in the Eth R&D Discord to stay up-to-date, or to join future Verkle Implementers Calls.
===
Total of five agenda items on the call
1. The relaunch of the Kaustinen testnet
2. Client updates
3. Conversion strategy
4. Post-state in the cryptographic proof
5. Verkle proof verification precompile
===
1. The relaunch of the Kaustinen testnet
Starting things off, @gballet shared a quick walkthrough of the new testnet: Kaustinen, which is now live and running. Thanks as well to @parithosh_j, @Gajpower, and @jasoriatanishq for their efforts.
See https://t.co/Vzbxxsho1i
We now also have a block explorer! Though it is still a WIP, it already will greatly improve our ability to follow along with what's going on in real-time. Currently the only CL client is Lodestar, and Lighthouse is coming soon behind
The Kaustinen testnet started off with Verkle from genesis, so there was no need to do any conversion. In the future, the plan will be to attempt a testnet that starts off with the Merkle Patricia Tree, and then does the conversion over to Verkle.
Few other changes live in Kaustinen: we now have the post-state in the block. Proofs are also being provided on Kaustinen. (But any client joining the testnet doesn’t need to be able to handle the proofs)
Using Guillaume's fuzzer, he was able to do some basic tests: e.g. make sure deploying contracts work, sending a bunch of transactions, and also tested withdrawals.
Next steps: work towards a future shadowfork 🔥
==
2. Client updates
condensed summary of client updates...
@gballet: besides the Kaustinen testnet launch, working on merging PRs into mainline Geth. Also doing some optimization and compatibility work with Ignacio.
@jasoriatanishq: @nethermind also joined the Kaustinen testnet, and it’s been stable for the past 5-6 days. @URozmej asked if there is the ability to have a stateless node. Tanishq: will create a new configuration for that and share instructions.
@ignaciohagopian: shared a document with some recent analytics from Kaustinen, using a modified Geth. This measures things like the average depth of the tree, witness size for each block, and proof generation/verification times. Super valuable tooling, though also important to note that this is all a bit experimental as we don't have much load on Kaustinen yet.
@kt2am1990, @DoctZed, and @gagadrupal: for Besu, working on some libraries as part of the Java implementation of Verkle. The plan is to integrate these libraries first into Bonsai in Besu. This may happen as soon as next week. At the same time, working on the modifications needed to join the testnet. Using the rust crypto libraries.
There was a question from Dragan on whether clients need to handle proofs to join the testnet.
@gballet: if it’s just about *following* the testnet, you don't need to verify the proofs. You only need to be able to receive an execution engine payload that contains them.
Daniel: for Nimbus, good progress, and substantial work on the primitives thanks to Advaita and Agnish. Using a library called Constantine from @m_ratsim. We have extended Constantine with the primitives needed for Banderwagon. Now in the process of integrating this library into Nimbus. Maybe 2-3 weeks away from seeing the first root commitment being computed correctly. Still a bit further away from joining the testnet.
==
3. Conversion Strategy
Next up, Guillaume brought up a question around the conversion strategy (how we migrate state from Merkle to Verkle), and whether the conversion should migrate N values per *block* or per *slot*
After a brief discussion, the consensus on the call was that per *block* is better. Since it is simpler and better in the case of reorgs.
==
4. Post-state in the cryptographic proof
@ignaciohagopian first shared a bit of context on the rationale for previously including the post-state values in the proof, which is to allow clients to know about the state difference that happened in a block, without needing to execute the block. But if they are not executing the block, they are actually trusting the new root of the tree.
Since you are trusting the root, you wouldn't gain anything by having the proof. After some discussion, realized it may have simply been due to prior confusion around the spec.
Regarding the post-state values: it is fine to use these values, as long as you are aware that you have to be getting these blocks from trusted or semi-trusted sources.
Ignacio has a working version of Geth with this updated, where the cryptographic proof only contains a proof for the pre-state, and *not* the post-state
@jasoriatanishq: Nethermind has an implemented version of a stateless client, and though there isn’t a spec, it is fairly straightforward. The idea is that you get the pre-state in the block, then generate a status tree from that pre-state, and you verify that the pre-state proof is correct. Now you have all the values you need for block execution. Instead of using the actual state from the database, you just use this state for block execution.
@gballet: Whether it’s stateful or stateless, it’s more of an implementation detail. It is up to the clients to decide what kind of store they want to have. And if they want to implement stateful or stateless it’s up to them.
@URozmej: the EVM remains the same.. the only change is that you get the state from the proofs rather than the actual database.
==
5. Verkle proof verification precompile
Lastly, Guillaume shared a new EIP which adds a precompile to verify Verkle proofs, so that dapps/bridges/etc can call this precompile and get the proof verified without having to understand the inner workings of Verkle. Goal is to reduce friction around the adoption of Verkle.
For a bit of extra context, in the future, it is likely that proofs are going to be distributed outside of consensus, and stateless clients will be able to choose their own proof format. This introduces an additional burden on applications. The precompile will make it easier for applications to support newer proof primitives without having to upgrade their contracts.
See the Eth Magicians thread to join the discussion: https://t.co/DhzAKfJyrE
Zahary added: the main use case for the precompile is bridges between different EVM chains.
==
6. Wrap up
Lastly, @gagadrupal shared a few quick things regarding Verkle in-circuits that he is currently researching. Spent some time looking into Zeth, which is a zkEVM project from RISC Zero. They showed that Keccak (perhaps unsurprisingly) takes a significant chunk of the resources for proving a block. But in Verkle, there is no Keccak, so that’s a small win already. And after the Verge it seems we will be in a better position to integrate with the latest ZK tech down the road.
That’s it! Next Verkle Implementers Call will be in 4 weeks, and from there we will be switching the call to a biweekly cadence ✌️