Plasma is a scaling technique where operations are moved off-chain into a secondary blockchain, where they can be performed faster and at lower cost.
The idea is based on “sidechains”, originally a proposal to scale bitcoin dating back to 2014. Plasma introduced a novel improvement: unlike on a sidechain, a user of a Plasma chain would always have a guarantee that they can withdraw their assets to main-chain, even if the operator of that Plasma chain tried to censor or steal from them.
Plasma research has made large strides since the paper release in August 2017, though the technology remains further from production than state channels. The year began with only a few teams actively working on Plasma, as the research community began to explore various tradeoff and design choices within a family of related techniques derived from the original paper.
The majority of these designs have focused on the simplest use case: payments. These designs include Plasma MVP (introduced by Vitalik in January 2018) and Plasma Cash (introduced by Vitalik & Karl in March, and the subject of continuing formal work). More recently, researchers have begun exploring zero-knowledge-proof based Plasma-like designs like the “Rollup” (introduced by Barry Whitehat in September).
Each of these can be augmented to mitigate their shortcomings, while minimizing tradeoffs. This unfortunately resulted in a “naming meme”, where each new tweak on an existing design was given a unique name, leading to significant confusion for anyone not deeply involved in the research community. Useful taxonomies that divide up the design space are a work in progress.
At the same time, research continued into expanding Plasma beyond payments. While this work continues, current consensus among researchers is that an optimized “full EVM” plasma (which could run any smart contract) is a complex challenge.
This broad exploration of a large design space has been productive for researchers, but practical implementations are still mostly theoretical or in early stages. One exception is the Plasma Cash implementation built by Loom, released in June 2018.
5. Zero Knowledge is coming
Over the past year the Ethereum developer community began to appreciate that new zero-knowledge technology will have a significant impact on blockchain technologies. Over the past 12 months, it has felt like every technical conversation in the Ethereum community takes the form of “well, we can do it this way for now, but of course once we have good zkSTARKs, it will be like this…”.
Most people in the crypto industry will have heard of zero-knowledge tech, most famously used in the privacy cryptocurrency Zcash. But zero-knowledge technology won’t just be used for privacy. It has important implications for many scalability techniques as well. Recent research and development into this technology (specifically, a class of zero-knowledge tech called zkSTARKs) may dramatically lower the computational cost required to use them in production, opening up new opportunities to integrate them with programmable blockchains like Ethereum.
In short, zero-knowledge proofs let us prove that some operation happened, without having to share the underlying data. If the verification of that proof can be done cheaply enough, then it could let Ethereum smart-contracts verify that an operation took place off-chain. This means that we could, for instance, conduct large numbers of operations off-chain, and then cheaply verify that they happened. Or, we could conduct intensive computation off-chain, and still have it verified on-chain.
2018 was the year that the full potential of zero-knowledge tech started to sink in. In January, Eli Ben-Sasson & his co-authors published their long-awaited paper on zkSTARKs. The Ethereum community began to work on how this technology can be used for scaling and in conjunction with other technologies, like Plasma. On the layer 1 side, developers made plans to ensure that ETH 2.0 has the requisite support for zkSTARKs, like STARK-friendly hash functions.
New zkSNARK libraries were released, like iden3’s snarkjs and circom, adding to existing libraries like Zokrates. In December 2018 a team at ETHSingapore built a zkSNARK “rollup” scaling proof of concept, later released on testnet as the (not technically Plasma) Plasma Ignis. BarryWhiteHat contributed critical work on using zkSNARKs on Ethereum. And Ben-Sasson & others launched Starkware, a company aimed at commercial applications of zkSTARKs, and received a $4mm grant from the Ethereum Foundation.
6. The Road to ETH 2.0
The history of ETH 2.0, aka Serenity, has featured false starts, dead ends and more false starts. But in 2018, the long-term Ethereum roadmap began to solidify.
In January 2018, the FFG testnet launched, though it suffered from networking issues that made it difficult to use. A few months later however, the research direction moved away from FFG, and towards a plan that would see Casper and Sharding implemented together. In Q2, consensus began to form around what is now the current plan.
Once the research vision was clear, it was possible to create a specification for what became known as “ETH 2.0”. This allowed many different engineering teams to begin implementing that specification into client software. At the end of 2018, there were at least 8 teams building clients for ETH 2.0. Recently, Ben Edgington also began a weekly newsletter which closely tracks the research and implementation of ETH 2.0.
While all roadmaps are subject to change and projections are uncertain, the beacon chain is expected to go live in 2019, with a beacon chain testnet scheduled to happen in the next few months. The beacon chain will allow ETH holders to choose to transfer their ETH to the beacon chain in order to earn rewards as a validator. However, that ETH cannot be transferred back to the “ETH 1” chain.
The next phase will include shards, which will be managed by the beacon chain. It’s also possible that the beacon chain will be used to finalize the current proof of work chain, somewhat similar to the way that FFG was once planned a year ago to be used for finalization.
While the roadmap has considerably firmed up, there are still unsolved problems in blockchain sharding. While the first few phases are relatively clear and there are no significant unsolved theoretical problems left, plenty of interesting research and implementation problems remain for future phases so that we get to a truly scalable layer 1 of Ethereum.
What did it all mean?
We warned you this post would be long. But it’s still not comprehensive. A lot more happened in the Ethereum ecosystem this year, including a few developments worth noting quickly here:
- Ethereum core developers came to rough consensus on a set of short-term upgrades to the current Ethereum protocol (“Ethereum 1.X”), while ETH 2.0 is under development
- Regulators across the world started paying attention to cryptocurrency, including securities regulators. Many jurisdictions are now in the process of deciding how digital assets, like the kind that can be created on Ethereum, are treated under law.
- Non-plasma sidechain technology, like POA network and Parity-bridge, launched into production
- UX made progress, like Universal Logins and meta transactions
- The Ethereum Foundation launched a grants program to fund critical work from across the community.
Should we view this year as a success, or a setback?
A strange thing about this moment in Ethereum’s history is that you will receive wildly different answers depending on someone’s frame of reference.
If your baseline is 2015–2016, you remember when Ethereum was still firmly an experiment, with virtually no users, developer tools, or even applications. The contrast with 2018 is startling. There are real applications now, live on mainnet, that provide real utility to their users — even if those user bases remain small. The thing we believed could happen, which may have once seemed impossible, is starting to happen, in bits and pieces.
But if your frame of reference is the hyped up narrative sold to you by ICO whitepapers and glossy keynote conferences, then it must be disappointing. Mass adoption not only has not yet arrived, but remains over the horizon. There are hard problems yet to be solved, and the technical progress zigs and zags, instead of following the comforting straight line of a tidy infographic roadmap.
Welcome to the real. There are experiments to be run, lessons to be learned, and hard problems to solve. Grab a shovel, and see you next year.