The launch of Ethereum 2.0 is estimated to begin this March with a Proof of Concept (PoC) of its core, the beacon chain, to go out this spring.
The beacon chain is the coordinator around which revolves much else. It tells stakers where to go, what shard to validate and so on.
A representative from Sigma Prime, which is working on an Ethereum 2.0 client called Lighthouse, said the timeline is difficult to estimate, but they’re hoping come March, April, we will be able to see clients talking to each other, pushing blocks around, having some interoperability.
A representative from Status, which is also working on a different Ethereum 2.0 client, concurred. Stating that they are aiming March for the Beacon chain.
He said they are working not just on the beacon chain, but also sharding “which is like a whole different world.” The timeline for sharding is much longer, he said, but for the beacon chain March may bring the first PoC, getting it out there to allow everyone to play around with it, validate assumptions, “give it to the community to start tearing it apart.”
He further added that during the first half of next year clients begin to talk to each other. That will pressure devs to get some of the fine points of the spec.
That will push the devs to deliver something, gives community something to play around and validate the ideas proposed.
Danny Ryan said that the Beacon chain will gradually be fleshed out. Vitalik Buterin later took a seat at the Council of Prague during the Status hackathon today.
First they’ll launch the beacon chain, he said, then data only sharded chains, execution added on later and then cross chain communication.
It will be rolled out in stages. Cross shards communication is not needed until state execution is launched. Every other stage can be done prior to it.
“The fundamental problems are figured out, more afraid of small stuff, bike shedding,” Buterin said according to notes.
There was a discussion involving randomness. Buterin said in Proof of Work (PoW), one guy makes a block, chooses where to build the block on top of and so on. In Proof of Stake (PoS), the same method can not be used as you can predict ranges during which you are a majority – can make short range attacks. Say within 20 blocks, 15 are mine, can perform a 5 or 10 blocks reversion.
There’s also selfish mining concerns and so on because the whole slot is one validator. Under those conditions, extreme randomness is very important.
The solution is to not use this approach of one validator at each height. Instead they’ve taken a validator slot, broken it up into 64 pieces. You can expect as long as more than 60% are honest, every single committee will be honest.
Ryan said there’s a separation between who is building blocks and a much larger group of validators dictating the path of fork choice.
Buterin said he’d like to prove base layer security under the assumption “random” hashes come from a textfile submitted by Danny Ryan.
They’ll be using randao, with a lot of work going on Verifiable Delay Functions (VDFs), but the protocol technically doesn’t depend on randomness being secure, Buterin said, because of shuffling mechanism – defense and depth component – randomness being completed is nice but not a pre-requisite.
There will be a voluntary migration. People will be able to deposit say 32 eth to the beacon chain, but it is not yet decided whether they will be able to send that back to the PoW chain. That would have benefits, but would add complexity.
“Parallel scaling,” Ryan calls it. The two chains will be running in parallel, linked by PoW blocks being used for finality in the beacon chain.
There may also be two roadmaps. The PoW chain, for example, could be upgraded to the Hybrid Casper chain which was deprecated according to Ryan who used to work on Hybrid Casper.
A roadmap for when to expect eWASM, the upgrade to EVM, wasn’t quite given. There will be a presentation on it at Devcon. eWASM is “the execution engine for eth 2.0 shardchains.” The EVM may be deprecated, turned into a contract. No decision made.
There will be a voluntary transition. People can choose to stay on PoW or go to the beacon. Same for dapps and so on. Solidity compiler adding eWASM at backend, Vyper has plans to do same thing. The transition therefore may be just a matter of recompiling for dapps, and for others just a matter of sending their eth to the beacon.
“Eth 2.0 is the same eth as the eth on the present chain,” Status rep said. Yet this all sounds like a huge endeavor, a migration, with it all to begin this estimated spring.
Incremental releases – most core devs have decided it makes sense, time for eth to grow up a bit and have more regular releasees. This will be demonstrated in Eth2 chain soon, Ryan said.
It sounds like there are parallel roadmaps too. Constantinople in the end might never make it, at least in any meaningful sense. That’s more PoW stuff, while now it appears attention might start turning towards the Beacon.
Sending 32eth means this is the launch of PoS. That has considerable implications for issuance as it is meant to be slashed down to just ~0.8 eth per block once PoS.
So how this all fits in with Constantinople, is not very clear. That is meant to go out in January. Beacon in March, but it’s just a Proof of Concept. Actual launch perhaps sometime later in 2019. At that point presumably we have PoW Istanbul to slash issuance to ~0.6, with the beacon giving 0.2 eth per block.
PoW miners have to upgrade because of the exponentially increasing difficulty which makes blocks more and more difficult to find. So they’d be earning a lot more by upgrading even at the much reduced issuance rate.
Eventually then presumably the PoW chain is deprecated in as far as the beacon chain doesn’t “care” much about it any longer, unlike initially.
In the meantime it looks like we’ll be seeing a lot of activity, with much of it to be fleshed out further at Devcon as coders are to be told about the great migration.