Source | What's New in Eth2
Author | Ben Edgington
recommended this week
Danny Ryan and Vitalik in the PEEPanEIP series gave a great presentation on what's coming with the Altair upgrade. Vitalik's presentation of "Light Client Sync" was an epiphany - I finally understood how the pieces fit together; it wasn't as complicated as I thought it would be.
beacon chain
beacon chain
The Beacon Chain has been doing well - nothing to say. New deposits to the Beacon Chain hit an all-time high in May - 26,681 new deposits.
The problem this week: 100 validators staked 64 ETH each. Everyone knows that a deposit requires 32 ETH, the excess is futile (no rewards will be earned), and it needs to be locked for about a year before the withdrawal is initiated. These validators first pledged 32 ETH each, and then in the same Eth1 block, the same validator added another 32 ETH to their existing deposit. These transactions are all done through Stakefish's batch deposit contract, and the two deposits pass through two calls with the same input data. It seems like a simple user error, but strangely, the second transaction with a gas price of zero was still included. According to Etherscan, they may have been submitted directly to the pool as they came from an F2Pool address.
first level title
Altair Testnet
Altair is a planned beacon chain upgrade that will take place in a few months.
As reported in regular meetings, the client team is implementing the alpha 7 specification for the Altair upgrade. The Teku team has set up a series of developer testnets, and other teams can try it out by synchronizing. It was a great learning opportunity, and it also brought some questions to the surface.
One of the issues we discussed at length in the meeting was when the sync committee should publish their signatures. The problem is that if another node receives the signature before it receives the relevant block, it will consider the signature invalid and discard it. Therefore, we either need to cache the receiver's signature for a short time, or delay the sender from broadcasting the signature.
As for Altair's plan, we hope to determine the fork slot of the beacon chain test network (Pyrmont or Prater) on July 1, the next meeting, in order to upgrade the main network in August.
merge
"merge"
"Merge" refers to combining the two chains of Eth1 and Eth2 to close the proof-of-work mechanism.
Justin Drake has gone through the Eth2/consensus layer portion of the "merged" spec and finished tidying it up.
In the regular merge meeting, we discussed the mechanism of replacing the soon-to-be-obsolete DIFFICULTY opcode on Eth1 with RANDAO of the beacon chain, because smart contracts are known to sometimes use DIFFICULTY as a source of randomness.
We also discussed the process for documenting the Eth1 changes required for the "merge". Basically, the standard EIP process will apply to Eth1 content changes that affect consensus.
pledge
pledge
first level title
Paraphrase
Reddit user Liberosist posted an insightful post "Understanding Ethereum's rollup-centric roadmap". This is about the updates and changes in what we call Ethereum 2.0 over the past year or so. I wrote about this evolving roadmap a few months ago.
We have been pushing for client diversity, so what about the combination of Nimbus + Hyperledger Besu? This article from QuickNode explains how to run Besu.
Avado published an article "Dangers of Staking on Exchange vs. Benefits of Staking on Avado", which I think is a good article. I could be wrong, but it seems that Avado's devices can currently only run the Prysm client. I will continue to pay attention to see if the arguments in the article about the dangers of staking on exchanges also apply to an ecosystem dominated by a single client.
Media and other resources
Attention, Beacon Book is about to start giving away! This book contains the perspectives of developers and others in various fields on the development of the beacon chain, and some very beautiful artwork.
Christine Kim's latest two Valid Points newsletters: 6.2 and 6.9 - Ethereum is rushing! 🚀
first level title
research work
MEV (miners/maximum extractable value) is a very hot topic. The Flashbots team did some research on MEV in Eth2 (ie the merged MEV). As a reminder, Vitalik's proposal to form a market for block builders continues to attract discussion.
(Translator's Note: Chinese versionandand"Preliminary Exploration of MEV in Eth2 (Part 2)")
In fact, in this week's implementers meeting, Protolambda explained that it could be of great benefit if we adopt a sharding design that resembles a marketplace between professional shard block builders and shard block proposers . He published a draft PR to explain the idea. Note that these proposals do not "solve" the MEV problem; they simply make MEV more equitably withdrawn by all, which helps avoid centralization tendencies. I think the block builder centralization problem is easier to solve.
secondary title
regular meeting
implementers meeting
Agenda
Agenda
Conference video (uploaded afterwards)
secondary title
"Merge" meeting
Agenda
Agenda
conference video
Protolambda's record
first level title
other news
first level title
write at the end...
Yesterday marked the fifth anniversary of the DAO hack - a pivotal and defining moment in the history of the early Ethereum blockchain. I really like Out of the Ether by Matt Leising, the whole book is about the DAO event. In terms of the scale of the incident, the DAO attackers took away 3.6 million ETH, worth more than $8 billion at today's currency prices. It is different from the usual DeFi fleece event.
At the time, I had been interested in Ethereum for a few months, although I hadn't really gotten into it yet. But the drama and excitement of the attack—technical issues, community reaction, white hats, failed soft fork design, and eventual hard fork—all captivated me. I'm completely hooked.
For a brief overview of the whole event, see Emin Gün Sirer's Twitter. (As many of you know, Gün and Phil Daian discovered the vulnerability in the code but mistakenly believed it was unexploitable.)
What an extraordinary time those were!
