In the past five days, I, our executive director Ming Chan, and several other members from the Ethereum team along with representatives from Ethereum-based initiatives and enterprises like Maker, String/Dfinity and Consensys have been traveling through China and Hong Kong. This involved a series of consecutive conferences and events, prominently showcasing the Blockchain Workshop in Hong Kong from Oct 11-13 and the “First Global Blockchain Summit” in Shanghai, organized by our associates at Wanxiang on Oct 15-16, all while we continued our regular responsibilities concerning research, development, planning for DEVcon, and administrative tasks.
Both events emerged as, in my somewhat biased view, exceedingly encouraging indicators of the expansion within the Ethereum ecosystem. During the Hong Kong gathering, on the inaugural day, we engaged in semi-private sessions addressing topics such as decentralized governance, identity and reputation, and regulatory compliance. Robin Hanson introduced the concept of prediction markets as a means of “low-cost micro-governance”: if instances arise where a legal or arbitration procedure is necessary to ultimately settle disputes, one could employ a prediction market to ascertain the outcome instead, escalating to the formal arbitration process only when necessary. Individuals with confidential insights are motivated to take part in the prediction market: the disputing parties, relevant third parties, and even companies like Google employing state-of-the-art machine learning models; in most scenarios, a prediction closely reflecting the result typically yielded by the foundational court or arbitration mechanism should be achievable at a minimal cost.
The subsequent two days entailed panel discussions revolving around both technical and philosophical subjects, with participants sharing insights on their own projects, alongside discussions regarding the future applications of blockchains in realms like reputation and the Internet of Things. Ethereum was referenced numerous times – not by us promoting ourselves, but by various individuals who regard Ethereum as a promising framework for their applications.
The Shanghai summit was equally remarkable. This was also a two-day program, with the initial day marked by a series of public addresses and panels, while the second day zeroed in on themed sessions; I personally participated in moderating the technical dialogues concerning consensus and scalability. This was quite a significant gathering, likely the largest since the Global Bitcoin Summit in Beijing, which had marked my first encounter with China over a year ago. Just as seen in the West during the same timeframe, the audience has evolved to be more mainstream: we observed representatives from major financial institutions, government officials, Intel, Huawei, and other traditional industry entities present, rather than solely employees from Bitcoin-specific (or even cryptocurrency-specific) projects.
There appears to be substantial enthusiasm among attendees regarding the adoption of blockchain technology for their specific needs – with ample interest directed towards Ethereum in particular. It remains accurate that there exist legal constraints, some quite evident and others ambiguous, impeding the potential uptake of Bitcoin or cryptocurrencies within China; however, there are currently no issues pertaining to blockchain technology itself. In the days preceding and following the conference, we dedicated time to engage with Wanxiang as well as other entities in China with whom we maintain regular communications, especially myself and Ming having our inaugural meeting with the remarkable team at http://ethfans.org/. We are still assessing how we can collaborate with Chinese individuals and businesses to maximize the usefulness of Ethereum in what has manifested as, based on purchasing power parity, the world’s largest economy, yet it is evident that the opportunities and potential are immense.
Research and Protocol Development
We have pinpointed four essential areas of investigation concerning the foundational protocol that we believe represent the principal milestones hindering the transition from the current state of Ethereum to a crypto-utopia:
- zk-SNARK integration: incorporating, whether via a new opcode or ideally by utilizing the EVM’s existing 256-bit modular arithmetic, the capability for Ethereum contracts to authenticate succinct zero-knowledge proofs. Considering that verification keys can be generated completely off-chain, this task is somewhat simpler than it appears, though once implemented, considerable infrastructural work will be necessary to enhance its utility. An initial objective will be to employ it for ultra-private coin mixing as well as for a privacy-preserving reputation (“here’s a proof indicating that I possess a score exceeding 250 according to your reputation assessment criteria and utilizing this dataset published to the blockchain, hence you should trust me, but I won’t disclose which parties you should trust me as”) and dual-party financial agreements, while the long-term aim will be the integration of Hawk on Ethereum; at the aforementioned conferences and outside of them, we have engaged in some fruitful discussions with a number of developers of these technologies on strategies for progressing towards this goal.
- Casper: The proof of stake mechanism for Ethereum, currently being developed by Vlad Zamfir with assistance from Lucius Greg Meredith, myself, and several others. The key components encompass by-block rather than by-chain consensus, and the notion of “economic consensus by bet” as a means of attaining de-facto finality with exponential speed compared to the linear speed characteristic of proof of work. The objective is to amalgamate a reduction in block intervals (I personally believe that a duration of 4 seconds will likely strike a good balance between safety and resource overconsumption and centralization risk; Vlad tends to adopt a more aggressive stance), finality, and significantly diminished energy consumption (and a general 10-100x lower price-of-anarchy). Presently, Vlad and Greg are engaged in the formal description and implementation of the non-economic elements of the algorithm to mathematically validate convergence properties,and the subsequent phase will involve enhancing the financial aspects.
- Scalability: utilizing a blend of sharding methodologies, random sampling, extensive employment of Merkle proofs, and asynchronous calls to elevate the potential transaction capacity from approximately 10-20 transactions per second to over 100,000 (or, if super-quadratic models are implemented, a theoretically limitless amount). The foundational principles behind scaling have been established for more than six months, and our research team is extremely assured that the overall strategy is sound; what remains is the specifics of exactly how to navigate optimal compromises that retain as much of Ethereum 1.0’s capabilities as possible in a scalable framework. A crucial design objective is to ensure that Ethereum 2.0 continues to encompass Ethereum 1.0; for instance, synchronous calls should still be feasible within a single shard, although cross-shard calls may necessitate asynchrony. I intend to revise my scalability paper to make it more focused, comprehensible, and updated with contemporary concepts over the upcoming month or so, although generally, scalability research is currently on the backburner until Casper is solidified.
- EVM enhancements: Martin Becze has taken the initiative to investigate WebAssembly as a potential candidate for an improved Ethereum virtual machine. WebAssembly shares numerous attributes with the EVM: the necessity to execute untrusted code, the requirement for exceptionally compact code, and the demand for several compatible implementations, although it does not include gas counting. WebAssembly can facilitate the creation of a just-in-time compiler for the EVM, significantly accelerating the Ethereum JS implementation, and it may also be utilized as a virtual machine alternative itself by introducing a transformation phase that incorporates a gas subtracting instruction to code before each jump action. Environment opcodes like BLOCKHASH, SSTORE, etc., can be abstracted by converting them into an ABI over a generic foreign function interface (for example, SSTORE(k, v) transforms to FFI(0x55 ++ k ++ v) where 0x55 acts as the marker representing SSTORE).
In addition, research inquiries persist on middleware “built on top of” Ethereum, encompassing on-chain services, decentralized governance, identity and reputation systems, random number generation, formal verification for Solidity (we have located an adept team in London that is eager to begin investigating the topic), and prediction market implementations, along with projects that may coexist “alongside” Ethereum (e.g., Whisper). However, the foundation is allowing the community to undertake much of this additional work since we have consciously adopted a strategy concentrating on the core (and doing so in a very streamlined manner).
Concerning Homestead, the Homestead milestone has always been somewhat subjective in its definition; unlike Frontier, which is distinguished by a grand and ceremonial launch of the blockchain, and Metropolis, marked by the release of Mist, Homestead was primarily intended to signify the transition from stating “Ethereum is unsafe” to “Ethereum is relatively safe.” Over the past two months, we have launched a wallet, alphas for state tree pruning in pyethereum and cpp-ethereum, headers-first syncing in Go, upgrades to Solidity, initial developments in the “light Ethereum sub-protocol” (LES), and are on the verge of releasing the 1.0 versions of cpp-ethereum and Mix.
Our internal goal has been to initiate Homestead when the Frontier network has successfully operated without significant issues for four weeks (the threshold for “serious issues” has been reduced from the Frontier definition of “consensus failure” to a broader definition encompassing something like “consensus failure OR individuals losing substantial amounts of money OR moderately serious network glitches”); we intend to disclose a more comprehensive strategy for Homestead shortly, but suffice it to say we are already mostly there.
DEVcon is still scheduled for November 9-13 in London, and we eagerly anticipate seeing everyone there!
