HackMDWhat's New in Eth2 - 3 December 2021
- The image file may be corrupted
- The server hosting the image is unavailable
- The image path is incorrect
- The image format is not supported
Edition 83 at eth2.news
⚠ Stakers, be sure to update your Eth1 nodes ASAP! ⚠
The Arrow Glacier upgrade takes place at block 13,773,000.
Etherscan has a handy countdown timer. As I write, it's showing Thu Dec 09 2021 18:34:47 GMT+0000. If your Eth1 node is not up to date, you may miss block proposals.
Top picks
If there is one treat you give yourself this week, make it Vitalik's updated roadmap diagram.
The Beacon Chain
The beacon chain passed its one-year anniversary on Wednesday. It's been a pretty epic year, and, to a very good approximation indeed, incident free.
I was planning to write an appreciation of the beacon chain's first year, but others have done the work already:
- Anthony Sassano has a nice write-up in The Daily Gwei, and
- Here are some stats from EthStaker.
And we're not only celebrating one year of the beacon chain, but also Lodestar's first mainnet block :tada:
Not everybody is partying, unfortunately. There have been a couple of slashings of genesis validators in the last few days, namely 1859 and 12697. No further info, but it's rather sad to get slashed after a whole year of successful participation. In other slightly tragic news, Lefteris stumbled upon one of several validators that have been offline since genesis. Here's another one.
Beacon chain spec things
Spec version 1.1.6 is out. This contains a fix for a very hard to exploit bug in the fork choice, and the proposer score-boosting defence against various attacks. These changes are not consensus-critical, so don't require a hard fork to be deployed in clients.
If you have noticed large increases in bandwidth and/or CPU usage in your beacon node recently, then it's probably due to a spec change that we are now reverting. This applies to all clients. You should hopefully see an improvement as the reversion gets deployed.
In other networking news, Danny has proposed a revamp of our attestation subnets now that we have some experience of running such a large p2p network.
Also from Danny, some ideas and proposals around how to implement infrastructure for weak subjectivity sync in practice.
The Merge
Tim Beiko published How The Merge Impacts Ethereum’s Application Layer to much acclaim, though I prefer my own suggested title for it. There is a version in Chinese.
The second community call on this topic took place today. I haven't seen video or notes yet, but keep an eye on the EF YouTube and something might appear in due course.
For reference, here's the Merge mainnet readiness tracker.
Kintsugi
Weekly Merge devnets are going pretty well. We put them up; we break them. Here's Barnabé behaving badly on Merge devnet-2. (Good news is that this is one of the things that spec v1.1.6 will fix.)
The plan is to start devnet-3 next week, and, all being well, a more public longer-lived Merge testnet around the 14th of December, which will be the Kintsugi testnet. Here's the tracker of clients' progress towards being able to stand up Kintsugi.
For lovers of detail, the following are the specific releases and PRs related to Kintsugi that I lifted from an internal ConsenSys presentation by Mikhail Kalinin (hope you don't mind, bro!):
- Kintsugi merge-devnet-2 has recently started
- Kintsugi spec v3 has been released
- Merge testing call has happened
As noted above, there was a Merge testing call last week. This prompted a flurry of activity from Marius of the Geth team who is rallying an army of volunteers to work on testing The Merge. Here are some pointers:
- Marius's Merge testing programme
- call to action
- over 300 people involved!
- Marius's own progress
For the brave, here are some instructions for joining devnet-2.
One of the main challenges for client teams right now is implementing "optimistic sync", which is a way to coordinate the consensus client and execution client when joining an existing network. There's a plan to specify the optimistic sync process formally. Meanwhile, here are some notes on it.
Staking
Rocket Pool is now fully live, and it has a pretty funky dashboard.
Lido has published its Q4/21 report on node and validator metrics.
MEV
If you've been paying attention, you'll be aware that maximal extracted value (MEV) post-Merge is becoming a big topic.
Flashbots is at the centre of trying to find ways to democratise access to MEV so that it is not all hoovered up by a few large staking pools, creating a massive centralising force for validators. To that end, they have suggested a network architecture that client teams can implement, and are coordinating a working group with a chunk list of stakeholders.
Part of the approach involves "proposer, builder separation" (PBS) in block production. In short, rather than building blocks themselves, validators propose blocks that specialist block builders (like Flashbots) build. The current approach is to implement this outside the consensus protocol, which involves a lot of trust assumptions. In future, there may be options to adopt something like this in-protocol, and ensure that it is fair and accessible to all.
Flashbots' MEV Roast 15 - PBS on Ethereum Roadmap was a session today covering all these topics. Unfortunately I missed it - I was planning to catch up via a recording, but I am not able to find one, despite some discussion of it on Twitter :cry: Will let you know if something turns up later.
One big concern around such PBS schemes is censorship resistance (especially while Flashbots has an effective monopoly on block production). Here are some thoughts from Vitalik on the topic.
Roadmap stuff
For the beacon chainiversary, Vitalik produced an updated Ethereum (2.0) roadmap diagram. Also available in Chinese, though they didn't attempt to translate the titles for each part :slightly_smiling_face:
The new roll-up centric roadmap is hungry for data. Vitalik has outlined a plan to gradually increase data availability via calldata expansion and staged sharding roll-out. This would increase total Ethereum ecosystem TPS sooner than waiting for the sharding big-bang in a year or two.
If we want to increase the amount of data flowing through the system, at some point we might want to start forgetting it too. This is where EIP-4444 comes in. There was asn impromptu AMA on history expiry with Vitalik on this last week. And Lightclients did an explainer of the background to it all.
The Great Explainers
Tim Beiko is everywhere this week. Here he is again, with Understanding the Transition to Proof of Stake. There are slides.
Guillaume and Dankrad explain how a practical implementation of Verkle trees might be implemented in clients. Verkle trees are one of the steps towards making Ethereum stateless.
Media and stuff
A couple of recent Bankless episodes are relevant. They're still on my playlist backlog, but I know they will both be great: (1) Decentralized Ethereum 2.0 Staking with Rocket Pool with Darren Langley and Dave Rugendyke, and (2) Layer Zero with Tim Beiko.
Coin Telegraph, Ethereum in full deflation mode as Eth2 merge gets closer:
Focus is now “exclusively on The Merge,” says Ethereum community manager".
Chainsafe had a bunch of talks at CSCON1. It's kind of hard to track down the videos right now as they are not indexed. But look out for:
- Alex Stokes on The Merge and beyond
- Lightclients on Ethereum accounts of the future
- And, it's him again, with a detailed update on the Ethereum roadmap
Regular Calls
Implementers
Call #77 took place on the 2nd of December.
- Agenda
- Video
- My quick notes.
All core devs
Call #127 took place on the 26th of November
Tim's latest All Core Devs update is an excellent read and summary.
Upcoming events
- December 9, 2021, 20:00 UTC: Post-Merge MEV Breakout Room
In other news
- Lighthouse Update 38
- Crushing Prysm's hasher by Potuz
- Danny's Finalized no. 32
And finally…
For my own birthday gift to the beacon chain, I published my fully revised and updated Altair annotated specification. If you pop over there, you'll get a strong hint as to what I'm planning to be working on for the next six months, and beyond. (Alongside the day job: never fear, Teku is my first love!). Wish me luck!
Learn More →
Learn More →
Advertising on this newsletter.
