Ethereum is working towards the biggest upgrade in the blockchain space since bitcoin’s invention a decade ago.
Unlike previous upgrades, this is not quite a hardfork, although there will be some of those. It is more of a complete transformation.
They plan to launch a brand new blockchain running on Proof of Stake with the initial Proof of Concept planned for March.
Called the beacon chain, this will eventually be the core of ethereum where all the blocks are found, produced, and all the shards as well as stakers are co-ordinated.
It will be incremental, starting with a transition period when the current Proof of Work (PoW) chain runs in parallel with the PoS Beacon chain, both linked by PoW blocks used for PoS finality.
That means there will be two parallel roadmaps, at least for some time. Some may keep working on the PoW chain and might perhaps turn that into Hybrid Casper, which was sort of finished. Most of the activity, however, and much of the attention will probably be on the beacon chain.
Once it goes live hopefully sometime next year, shard chains will then be added as a data layer, with the execution layer to follow. Cross links between shards are then to be implemented, with the sharding of the shards potentially following.
The roadmap therefore is unlikely to be finished for the next five years at the very least as quadratic sharding is complex. Improvements, however, should in the meantime add considerable capacity, with it perhaps beginning to be felt starting in 2020.
In 2015, Buterin said at the Council of Prague yesterday, ethereum’s roadmap was laid out in stages, starting with Frontier. That was meant to be the experimental first version of eth, with Homestead meant to be the stable version. Then there was Metropolis, which is still kind of being finalized in the Constantinople upgrade expected in January.
The final stage is Serenity, which was more vague at the time but it included the launch of PoS, with ideas regarding sharding then developing.
5/8th of the 2015 vision has been implemented, Buterin said. Before 2017 it was less clear what PoS would look like, he added. There was CasperFFG, which was a simplification of PoS, but then ideas formed around sharding.
The initial approach was a slow roadmap of first HybridPoS, then full PoS. Buterin said he was sad to see it go as he wrote much of it, but it make more sense to do the beacon chain PoS architecture.
That suggests this is very much full PoS, although it will operate sort of in parallel to the PoW chain initially, but it sounds like that PoW chain can somewhat easily be discarded eventually.
Much of the ideas around this transformation to a scalable blockchain are shared on ethresearch, a forum set up for more abstract discussions. That’s where the beacon chain and all the rest was discussed and continues to be discussed.
Now, however, this is starting to take a more solid form, so some discussion is moving to opening issues and pull requests on Github.
They have moved past editing the whole spec every week, with changes now much more incremental. So it has gone out of hackmd and onto Github.
Meaning this is now at the implementation stage, with a number of clients working on it based on the specification.
Much more detail will now be provided at Devcon. which opens tomorrow but starts properly on the tenth birthday of bitcoin this Halloween.
We are hazy, for example, on just how this parallel roadmap will work. We can have a fairly good guess, but we’d want to know whether there will be coordination between the beacon chain teams, the sharding teams, and the more PoW current clients like Geth and Parity.
Afri Schoedon of Parity said he was not up to date and that Parity was behind other teams. Geth likewise is probably not working on Ethereum 2.0.
So for now it is perhaps a bit too early for such fine details as to whether we will get PoS clients in addition to PoW clients. The former are not quite competing, but obviously prestige comes with market share. So we’ll see which one the market eventually favors.
Just as we’ll see whether eth’s issuance is to be reduced to circa 0.8 eth next year. Down from the planned 2eth per block, bringing it lower than bitcoin prior to its halving in 2020.
Likewise it is probably far too early to get a feel for the new chain as far as end-users are concerned. If they move eth there, is it as if nothing has happened or are there different address formats and so on.
For dapps and smart contracts, they’ll probably want to know whether they can keep running on the new chain, and if they want to move, whether it is a lot of work or just a matter of recompiling.
Our guess is that there’s no difference from the end-users perspective and for dapps they said a move to eWASM is just a matter of recompiling, but it isn’t quite clear when eWASM is planned for launch.
With much thus seemingly in flux as it appears a lot has been going on and a lot will be going on.
People will have to voluntarily move to the new chain, as will dapps. They can stay where they are if they want, but hopefully the new chain will be a lot better, especially for stakers.
Then as sharding is rolled out it will be a lot better for all due to considerable capacity. Meaning that perhaps patience will pay off as the world may get a far nicer chain, with considerable scalability and with what sounds like full stake.