What's New in Eth2 - 02 October 2020

Ben Edgington (Eth2 at ConsenSys — all views expressed are my own)

Edition 53 at eth2.news - the second anniversary edition.

I am blown away by such wonderfully generous support from you all for my Gitcoin Grant during this matching round (now closed). You have given me renewed encouragement to press on with the Annotated Spec! Beacon chain notes finished by launch???

Top picks

Danny's eth2 quick updates no. 17 and no. 18 are brief, but introduce and say farewell to Spadina, and announce the Zinken network, among other things.

But this is my favourite read this week: A rollup-centric ethereum roadmap. Much to say about this, and I'll say some of it below.

Phase 0: The beacon chain

We're almost there! A v1.0 release candidate beacon chain spec is in the works. I believe the deposit contract deployment to be days, not weeks, away. And beacon chain genesis weeks, not months, after that.

Testnets

Medalla

The Medalla testnet continues to run smoothly. As of writing, there are 66275 active validators, with around a 75% participation rate. It's reached slot 426660 in epoch 13333, which means that it has now been running for almost 60 days.

Afri did some comparative sync speed benchmarking between the clients. Yeah, I know Teku looks lame. We know what's slowing us down, and we have an improved sync coming. But we've been focusing instead on implementing the weak subjectivity sync from a provided state, which sidesteps the whole issue. Long syncs from Genesis are risky in Proof of Stake, and for safety and user experience reasons will not be necessary in future.

Spadina

The Spadina testnet came and went, just as it was supposed to. This was intended to be simply a dress-rehearsal for the beacon chain launch. Unfortunately it was a bit bumpy, basically due to some lack of preparation (Prysm's postmortem). We learned some lessons, Danny told us all off, and he's making us do it all again with Zinken.

There was a very nice launch party for Spadina, with special guest Camila Russo of Infinite Machine fame, alongside all the usual suspects.

Zinken

Zinken is the re-run of Spadina. This time we are allowing a longer period between the calculation of the genesis state and the start of the chain (it was 2 days for Spadina and Medalla, will be 4 days for Zinken, and 7 for mainnet). This will allow everyone time to get their ducks in a row.

The Launchpad for Zinken is available. Make your Goerli Ether deposits before October 8th, noon UTC if you'd like to run a validator. But please treat it as you would the mainnet launch itself: watch for announcements from teams; keep your client up to date; keep your client running.

Tooling

The Eth2 deposit CLI has been audited, and the findings are being addressed. This is a tool that works with the Launchpad to generate validator keys and register them with the deposit contract. We (Teku) will be recommending that stakers use only audited tools for this, and we are planning to remove our current convenience deposit tools.

Our excellent Automated Verification team in PegaSys has performed a full formal verification of the incremental Merkle tree algorithm used in the Eth2 deposit contract. This goes beyond the work done by Runtime Verification which was a partially mechanised proof.

The good folk at Attestant have been busy as ever. chaind is a process that reads information from an Ethereum 2 beacon node and stores it in a database for reporting and analysis. It is particularly useful for the Medalla data challenge (entries close on October 20th!). And Vouch is a very interesting approach to running a validator. Rather than having to select a single team's validator client for their staking infrastructure, Attestant wrote their own validator client that can talk to a variety of different beacon nodes. This allows them to optimise for a range of metrics by working with the most effective node at any given time.

Secret Shared Validators

Secret shared validators are an architecture for distributing staking nodes across multiple hardware instances by using multi-party computation. Last week, the first SSV went live on the Medalla testnet. It's a little up and down currently, but this is an amazing step forward. There are four validator clients using a threshold key, coming to consensus on which attestation to sign via Istanbul BFT.

There was a Secret Shared Validators community call today. I'll link the recording when it is available. There's also a discussion channel on the Ethereum Research Discord.

Rollups and Eth2 Phases 0, 1, 1.5, and 2

Vitalik pulled together a bunch of threads in his article today, A rollup-centric ethereum roadmap, and wove them into a compelling vision.

Rollups are a way of reducing the load on a blockchain. They allow the expensive heavy-lifting (such as maintaining a record of account balances, and, in some cases, general computation) to be done off-chain (at "Layer 2"), while only a compressed record of the transactions goes on-chain, or a proof of the ledger's correctness. By doing this, rollups go some way towards delivering the stateless Ethereum that we need so badly. Keeping rollup operators trustless and honest is done by means of zero-knowledge proofs (zkRollups) or by challenging them with fraud proofs (optimistic rollups). You can use a zkRollup today on Gitcoin Grants!

So, how does this all fit with Eth2? Briefly:

• Phase 0, Proof of Stake: no impact.
• Phase 1, Data sharding: rollups rely on data availability to deliver their high transaction throughputs. What is Phase 1 if not a massive data-availability engine? Potentially the Eth1 chain could be a light client of Eth2 Phase 1 (i.e. Eth1 could see and verify transactions on the Eth2 chain). In this way, rollups running on Eth1 but viewing Eth2 could deliver huge transaction throughput.
• Phase 1.5, Eth1/Eth2 merger: If most of the activity on Ethereum is taking place in rollups rather than on-chain, then this migration potentially becomes much easier and less disruptive.
• Phase 2, sharded execution environments: well, rollups are shards, kinda. Maybe we don't need Phase 2 any longer. This simplifies the base protocol, which is always good.

None of this will come as a surprise to anyone who has been paying attention, but it is interesting to reflect on how the context has changed during the 2.5 years we've been planning Eth2. The emergence of rollups is a happy consequence of the length of time it is taking to get Eth2 done. If we had "solved" gas costs already, this technology would not have emerged, and the Ethereum roadmap would be immeasurably weaker. Truly effective scaling happens at both Layer 1 and Layer 2.

The Great Explainers

Need some help getting into Eth2 staking? Experienced, and happy to help others out? The wonderful Ethstaker community is organising community mentorship groups to connect people with others who can support them. I love this.

More generally, I do really recommend the Ethstaker Reddit. This is a lovely community, very keen on exploring staking as individuals. There's a wealth of helpful articles there, such as A slightly updated look at hardware for staking, eth2-docker walkthrough … "easy" node setup, and of course, the Eth2 StudyMaster Programme.

Tetranode has been live-tweeting the experience of setting up Eth2 staking on Windows. If you want to make it easier for yourself, check out How to stake for Ethereum 2.0 with DappNode.

Media

Chase Wright prompted an interesting discussion with a Twitter thread, Why I won't be staking ETH at launch, the main thrust being the he considers Eth2 unready for launch. He gets some things wrong, some things right, and others are a matter of opinion and risk/reward appetite. Danny Ryan produced a characteristically level-headed and informative response. But the point is well-made that individual staking in Phase 0 will not be for everyone. There is definitely a hurdle to jump in terms of technical knowledge, commitment, and risk tolerance. You need to do your own research and make up your own minds. Third-party staking services will reduce the risks, at a cost of reduced rewards and reduced network decentralisation. Otherwise feel free to sit things out until you feel confident. This is not some ludicrous DeFi Ponzi where only the first get all the rewards.

Research

Vitalik is still looking for answers around ideal vector commitments for "an optimal scalable blockchain".

Also on ethresear.ch:

• Eth2 attack via time servers outlines a griefing attack against validators via NTP manipulation. Alex Vlasov of PegaSys has written extensively about robust, decentralised approaches to clock synchronisation.

Regular Calls

Implementers

Call #49 took place on the 1st of October.

• Agenda
• Video
• My quick notes and from Mamy.

The call was pretty brief; not too much to note, mainly just updates.

In other news

• Client team update from Prysm.
• A fuzz testing update from Sigma Prime. Critical errors found in the Blst library, Teku now integrated, and much, much more!
• The Ether has a Validator board for claims about running validators on Eth2. As far as I can tell, The Ether is a platform for gameifying debate. Share your opinions, earn XP.
• Ethstaker is running a very quick survey of people's intentions around staking. If you are planning to run your own node for staking on Eth2, please check it out.
• ETHOnline is happening! I missed Danny's talk today (I came in right at the end when he was promoting What's New in Eth2 ), and Vitalik is on after my bedtime, but I guess the talks will appear here in due course. October 23rd is all around The Future of Ethereum, and I anticipate some great Eth2 content.

And finally…

It's two years since the very first edition of What's New in Eth2 hit the stands. These things that you start so casually have an unnerving way of taking over your life - beware!

HackMD tells me that we're getting somewhere between three and five thousand views per edition. I don't know how accurate that is, but

I never imagined that my late night ramblings might interest so many people. The tone has changed over time: it's less overtly technical now; I am trying to do a little more explaining, and am daring to share some more opinions. It's a more fun way to write.

I haven't sought or received much substantive feedback on WNIE2 over these years. But the feedback I have received has been super helpful, and guided its course for the better.

With that in mind, I've made a super basic feedback form, which is totally anonymous. If you are happy with what you're getting, then just keep reading, there's no need to let me know. But if you have any suggestions or constructive criticism, I'd love to hear it. Thanks in advance!

---

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

 We also have an RSS feed.