Ethereum 2.0 developers in at least two client implementations are finalizing preparations for a testnet launch of the brand new Proof of Stake (PoS) Beacon Chain.
“The testnet launch is getting pretty close,” says Dustin Brody, a Status developer working on Nimbus, one of the new eth 2.0 client.
— Nimbus (@ethnimbus) March 13, 2019
Yuriy Glukhov, another developer at Nimbus, says “we’re going to launch [the testnet] soon, as far as I’m aware.” Explaining what’s going on in the above tweet, Glukhov said:
“The simulation is running a few beacon nodes with validators attached to them (some validators are missing on purpose to test the reaction of the majority). The validators propose the blocks and attest to them, the nodes communicate to each other.”
That means the Nimbus implementation has now reached the stage where devs can run an early version of the node after months of hard work with the video showing benchmarking and a sort of lab-setting test that shows it all works fine in a sole laptop/computer.
Another implementation at the advanced stages is Sigma Prime’s new Lighthouse eth2.0 client. Speaking to Trustnodes, Age Manning, co-founder of Sigma Prime, said:
“We’re benchmarking the spec with our implementation and adding various optimisations to speed things along. We’re also in the process of implementing the phase 0 wire spec (which is not fully agreed on yet).
We’re hoping to have a single client testnet with the phase 0 wire spec soon. Over the next few weeks we’ll be focusing on the networking and syncing aspect of the client…
All the clients will have varying degrees of client/networking/syncing infra structure (especially because we don’t yet have a guideline for how nodes should communicate)…
For lighthouse, all the spec is implemented and partially optimised. We have been running benchmarks for 4M validators to ensure the implementation can support such processing.
The remaining section of our client needed to form a testnet is simply the networking and syncing. We could simply open tcp ports and develop custom messages to facilitate this. However we are planning on going down the path of implementing the rust-libp2p infrastructure to fit the phase 0 spec as much as possible.”
So the testnet may now launch in a few weeks with it initially being a single client testnet. That means different people/computers running the same node implementation. Then there will be a cross-client testnet to see they in sync, audits of the code by independent third parties, and the live launch by perhaps autumn or winter according to estimates.
This is a very new chain, but connected to the Proof of Work (PoW) chain. There’s a recommended video that explains how it works by Justin Drake, a researcher that has been working on this for months.
The main thing to understand is that the Beacon co-ordinates all the stakers/validators in the different shards. In addition, initially it also kind of validates PoW blocks. Explaining how the Beacon knows that 32eth has been locked in the PoW smart contract, Manning said:
“The Beacon chain has a PoS consensus mechanism, meaning that validators vote on beacon blocks. But they can also vote on the latest Eth1 chain block hash (at the same time). So essentially, we use the PoS mechanism that is already there to agree on the Eth1 chain blocks in a decentralized way.”
That presumably translates to stakers have to run a PoW eth node in addition to the Beacon node. Since the eth node obviously knows what’s going on in all the smart contracts, it also knows whether the 32 eth has been deposited, with the Beacon then having a “Deposit” object in the beacon chain.
“Other chains definitely can agree on and include Eth1 blocks if their consensus or processing layer allow it,” Manning says, so potentially creating an interoperability method of connecting different chains.
Once this is live, any etherean with 32eth can then deposit it to the smart contract and become part of the consensus process with the role very similar to that of a miner.
In return for securing the network, stakers get between 5%-8% of their staked amount per year, making eth sort of a dividend bearing asset and thus with an “intrinsic” return on investment.
Once it is launched, it will reduce new supply to 0.82 eth per block from the current ◊2, with the shards then to go out in a year or two according to very general estimates.