The Stateless Ethereum Summit
It would be an unwise endeavor to attempt to deliver a representative or impartial overview immediately following this week in Paris — both I and others in attendance will be spending the upcoming weeks refining our insights and preparing for the year ahead.
However, for you, dear reader, who experienced the Paris FOMO and have been eagerly anticipating an update, I will share my personal and incomplete compilation of high-level observations, decisions, and outcomes from the inaugural Stateless Ethereum Summit.
What was the experience like?
The summit lasted two days, featuring a minimal structure that began with the entire group convening to discuss significant or pressing subjects, followed by break-out sessions with two or three simultaneous conversations. With around 30 participants in total, the group sizes were nearly ideal for facilitating both in-depth discussions and casual Q&A. Additionally, it served as an opportunity to associate faces with usernames and connect on a more personal level with everyone involved.
For many attendees, including myself, the central outcome of the summit was a “leveling up” in our comprehension of the challenges that require solutions and the suggested approaches. The small collective of individuals spearheading this initiative (Piper, Alexey, and their respective teams) provided the rest of us with valuable whiteboard time to catch up and pose all the minor inquiries that we hesitated to raise in a forum post.
I emphasize this because one of the principal objectives of this gathering was to present more clearly both the opportunities and obstacles of the work to be undertaken. The clearer this work can be articulated to everyone interested, the easier it will be for them to join the initiative and contribute. In this regard, I would assert that the summit was already a significant success, successfully “hooking” some individuals who had been observing from the sidelines until now.
What topics were explored?
Well, almost everything. With just one pair of ears, I absorbed many subjects from the tech tree being discussed in detail, and as mentioned earlier, this summit was fundamentally about uniting to agree on a simple shared vision for Stateless Ethereum. What is the fundamental issue we are addressing? What is the first attainable milestone to pursue? Is it worthwhile to explore a zero-knowledge framework for historical witnesses?
Here’s what I perceive were the primary subjects:
- Syncing primitives
- Transition to Binary trie
- EVM
- Data delivery in the stateless context
- The draft witness specification
Alexey astutely noted that the purpose of this summit was to accomplish all the tasks that couldn’t be achieved online, and to reserve tasks that can be done remotely for when we reconvene virtually. One aspect that functions significantly better in person than online is disagreement, along with relatively swift decision-making on intricate issues. Consequently, in addition to the general recap and knowledge-sharing regarding the core topics of discussion, there was a focus on utilizing the time to advocate for or against decisions that needed to be reached, such as what to prioritize or which new tools are necessary prior to initiating any work at all. Most crucially, this summit provided a chance to refine and better delineate the scope of this work while collectively gaining a clearer understanding of what success looks like from various perspectives.
What conclusions were drawn? What’s new?
Once again, and I cannot emphasize this enough: This is merely my personal brain-dump summarizing the summit’s outcomes. I have yet to review my notes and recordings. However, here are my takeaways, in no specific sequence. These represent new insights that emerged during the weekend’s discussions and will influence future directions.
-
Sync, and more specifically the primitive getNodeData is the essential aspect that must evolve to advance this stateless endeavor. It is a necessary modification that must occur before the transition to a binary trie can take place, and it will demand coordination among all client teams. Felix from the geth team facilitated a highly productive dialogue on sync, and it became encouragingly evident that most alternative sync proposals appear to be addressing the same objective from different perspectives. Enhancing sync will facilitate a smoother transition to a binary trie as well.
-
While it was previously believed that the reliable transition strategy to a binary trie would necessitate a brief suspension of the chain and a recalculation of a new binary state, the latest thoughts indicate that the transition can be carried out without network disruption with adequate client cooperation.
-
The intentions and concepts about establishing a full-fledged Ethereum-specific data delivery network for state have been significantly altered due to a combination of newly acquired insights. Firstly, experts provided clarity on just how challenging it would be to construct such a system. Secondly, this network can be incrementally developed from enhancing sync, and a simpler version (serving only headers, transactions, and receipts, for instance) would bring immediate value and could be improved later.
-
EVM modifications are the most intricate, and there was no definitive conclusion or agreement regarding what EVM alterations will be necessary for stateless compatibility. The challenge here is that many proposals under evaluation actually do more than what is strictly needed for statelessness, raising the question of assessing the value/complexity/effort for those additional enhancements. It’s worth mentioning that some gas operations are anticipated to increase in cost, regardless, but nothing is definitively settled concerning the EVM, and we will not determine the optimal direction until we acquire more data.
-
WE MUST BUILD ADDITIONAL PYLONS — This is a quirky way of stating that some of the forthcoming efforts will concentrate on enhancing the productivity and outcomes of the work itself. This meta-work divides into two categories: Tools that simplify data collection and analysis, and resources to aid others in contributing more effectively, such as stateless-specific documentation for new researchers joining the community. That said, I believe there exists significant disagreement on how much effort should be allocated in the short term towards tool development, and which tools are of highest priority. In the upcoming weeks, we intend to revise the tech tree and develop it into something more representative of the initiative that Stateless Ethereum has evolved into. This will serve both to assist the community in keeping track of everything and to enable interested newcomers to contribute more effectively.
As always, if you have inquiries, suggestions for new topics, or wish to get involved in stateless Ethereum research, feel free to introduce yourself on ethresear.ch, and/or contact @gichiba or @JHancock on Twitter.
