Try   HackMD

What's New in Eth2 - 17 December 2021

Image Not Showing Possible Reasons
  • The image file may be corrupted
  • The server hosting the image is unavailable
  • The image path is incorrect
  • The image format is not supported
Learn More →
Ben Edgington (Eth2 at ConsenSys — all views expressed are my own)

Edition 84 at eth2.news

Top pick

The surprise announcement of the week was the Ethereum Foundation's client incentive program(me). The programme provides beacon chain stakes to both Eth2 and Eth1 client teams. Initially, the EF retains the withdrawal keys for the stakes, but over a few years post-Merge the keys will be handed over, and teams can then do as they please with the stakes and the accumulated staking rewards. I love the incentive alignments encapsulated in this mechanism - it really functions nicely to keep teams in the game for the long run. It's kind of an open secret (oddly not mentioned in the announcement, though the last paragraph does say "to finally share") that the four genesis Eth2 clients have all been running stakes under this programme for over a year now.

The Beacon Chain

Tooling

Miga Labs has officially released Ant Crawler, the first version of its Armiarma beacon chain network crawler. It is the engine behind their Eth2 network dashboard.

Chainsafe has a crawler and dashboard too. As ever, we can see that this network crawling business is not an exact science; results differ significantly between the two. And Stereum's client diversity watch is completely different again. Miga Labs' write-up discusses some of the complexities involved.

Also note that all of these count nodes of each client, which is very likely not representative of the amount of stake (validators) managed by each client, the more important metric.

In other tooling news, the Wagyu Key Gen tool has been audited. Regular readers know that I love a good audit report, so here it is. (Hey Mikerah!

Image Not Showing Possible Reasons
  • The image file may be corrupted
  • The server hosting the image is unavailable
  • The image path is incorrect
  • The image format is not supported
Learn More →
, been a while.)

Light client corner

A couple of things popped up around light clients. The beginnings of support for light clients were introduced with the Altair upgrade, via sync committees.

  • Lodestar released the first beacon chain light client. You can try it out.
  • Not satisfied with a light client? How about a light node. Nimbus is looking at extending the light client protocol to rely only on a 1-of-N honesty assumption among one's peers. There's a good collection of light client resources at the end.

The Merge

Kintsugi

The Kintsugi testnet is live! This is the first significant public Merge testnet. You can check the network status, send transactions, run a pair of clients (execution plus consensus), and generally mess about trying to break things. All the details about participation are at kintsugi.themerge.dev. Also see Remy Roy's guide for Geth+Lighthouse (and a walk through video with Superphiz). Other clients are available

Image Not Showing Possible Reasons
  • The image file may be corrupted
  • The server hosting the image is unavailable
  • The image path is incorrect
  • The image format is not supported
Learn More →
: here's a list of Kintsugi-ready client configs that may or may not be up to date.

Huge props to Pari for setting up most of this. Check out his console array showing the various client combinations going through the merge process itself. (Teku is the second column in all three rows.)

It's a "Great weekend for testing #Eth2"! And Marius is still looking for volunteers for all kinds of testing.

Syncing

The main engineering issue remaining to be solved before the Merge relates to how the combined clients will sync. It turns out that efficiently syncing up two semi-independent clients (Eth1 and Eth2), each managing its own world state, is not straightforward, with particular hazards involving the Merge block itself. Under the Merge architecture, the consensus client takes the lead, but it cannot validate every execution payload while the execution client is syncing. Meanwhile the execution side is doing some sort of snap sync to build its world state at some recent time in the past it's complicated.

The current leading option is "optimistic sync", but this is proving tricky to implement in clients, with some nasty corner cases. An alternative is for the consensus side to take advantage of the new Altair capabilities and perform a light client sync before bringing the execution side up to date. This comes with its own trade-offs, as discussed on this week's devs call.

Fundamentally, this is an engineering challenge, and we'll reach a workable solution soon enough. The Ethereum network has had a variety of different sync types and formats over the years, and no doubt we will continue to iterate on this for the foreseeable future.

MEV

Post-Merge Maximal Extractable Value continues to be a huge topic. Flashbots is coordinating a working group that many stakeholders (including ConsenSys and Teku separately) are participating in.

For a very good overview and rationale for Flashbots' approach, see Robert Miller's Why Building the Most Profitable Block is Important.

A breakout room call took place on the the topic last week, which included a walk through of Flashbots' MEV_Boost proposal, and a demo of it in action, with Q&A.

Other Merge news

I was a bit surprised to learn that Gnosis has launched their own version of the beacon chain. You can stake and earn $GNO tokens using versions of Lighthouse or Prysm. The stated idea is that it can act as a kind of canary for the Merge, implementing updates a week or so ahead of the beacon chain, and eventually merging with xDai (now known as Gnosis Chain) for its execution layer. Not everyone thinks that this is a great idea (IST is "irregular state transition" - the point being that xDai is more than just a testnet, so a degree of care is needed).

Meanwhile, the great Merge beacon chain upgrade naming exercise continues. A shortlist has been assembled, and the final poll will be taken on a Zoom call on Monday.

and beyond

All eyes are on The Merge right now, but we must not forget that there remains plenty of work to do afterwards. In case Vitalik's roadmap were not enough, Tim Beiko posted the list of open issues on the Eth1 GitHub as part of his All Core Devs call notes. Never mind a few odds and ends like withdrawals and sharding that we need to deliver on the Eth2 side as well

Image Not Showing Possible Reasons
  • The image file may be corrupted
  • The server hosting the image is unavailable
  • The image path is incorrect
  • The image format is not supported
Learn More →

Where does it all end? Vitalik has ideas.

Staking

Lido continues to move towards ever greater transparency with a new operator performance dashboard. It is also open to non-Lido operators - it would be outstanding to gain a properly comprehensive view across staking services.

Regular Calls

Implementers

Call #78 took place on the 16th of December. Unfortunately, due to a technical fail, the meeting was not streamed or recorded, so my brief notes remain the only record of the call. I skipped over some fairly detailed discussion on optimistic sync, but much of it has been re-hashed on the Eth R&D Discord in the merge-general channel.

Mostly updates on the Kintsugi testnet, discussion of optimistic sync and light client sync, and regular client updates.

We will skip the next call that was due between Christmas and New Year, and meet again on the 13th of January. There is an All Core Devs call in three weeks if anything more urgent needs to be discussed.

Merge Community Call

Merge Community Call #2 took place two weeks ago:

Upcoming events

In other news

Neither of these are specifically Eth2 related, but I wanted to highlight them anyway. Two massive resources for newcomers:

  1. A collaborative effort to create a guide to joining the Ethereum ecosystem.
  2. Patrick McCorry is launching "a free 8 week course on cryptocurrencies with a special focus on layer-2 protocols". This will be outstanding. I think sign-ups are still open.

Other bits and pieces:

And finally

I'll probably not be doing an issue in two weeks, it being New Year's Eve and all. Not that I have anywhere to be, but you'll all probably have better things to be doing (like testing Kintsugi). I'll pick up again on the 6th of January, or maybe the 13th to keep in step with the Eth2 devs calls. Depends how much news there is

Image Not Showing Possible Reasons
  • The image file may be corrupted
  • The server hosting the image is unavailable
  • The image path is incorrect
  • The image format is not supported
Learn More →

Meanwhile, I wish you all a wonderful Christmas, and a Happy New Year. See you on the other side!


Follow me on Twitter to hear when the next edition is out 🙌.

 We also have an RSS feed.

Advertising on this newsletter.

Give Feedback.