Nov 1st, 2017- Cancun, Mexico
Jump to Summary
Including Casper for Proof of Stake, Plasma & Sharding for scalability, zkSNARKs for privacy.
(and much more...)
A reminder about the basic principles of the Ethereum network.
A discussion on approaches to parallel processing for Ethereum transactions.
Robert discussed 3 different types of Ethereum clients:
Full Node — downloads all block headers and bodies, runs through consensus algorithms, verifies all transactions
Light Client — Just verifies block headers, but no transactions
Thin Client — Verifies block headers, but does not validate state transitions.
Entertaining talk from Yoichi on the voting mechanisms proposed for Casper, the Proof of Stake proposal for Ethereum.
Interesting talk from Lefteris on Sikorka — a framework for “Proof of Presence” in Ethereum.
Proof of Presence in this case means proving an entity was at a specific location in space at a specific point in time. Applications for this could include proving you were in a store at a given time to receive loyalty points for example.
Alex laid out a case that testing & auditing Solidity is expensive and error prone, and presented the Julia language as a solution for this. Julia is a simpler language designed to increase confidence in code testing and auditability.
Piper discussed package management for Smart Contracts — something very much needed in the space
Great Talk on the methods and importance of iterative incentive design & test, to produce good outcomes.
The voting mechanisms within Ethereum’s Casper are a great first example of this kind of balanced incentives structure.
A correct-by-construction blockchain
Discussion on the design implications of Casper.
TrueBit virtual machine — a scalability solution for compute in the Ethereum network.
Jason put forth the idea that incentives for miners need to balance both correctness and participation where in some cases miners can be incentivized to accept incorrect blocks — the so-called “Verifier’s Dilemma”
The idea for Plasma is essentially an N-level hierarchy of blockchains. One main parent blockchain (the Ethereum Mainnet), and any number of child chains branching off from it, with recursive child branching as needed. Each Plasma chain/branch can define its own consensus algorithm.
Jacob discusses zkSNARKs and off-chain processing. zkSNARKs = Zero Knowledge Succinct Non-Interactive ARguments of Knowledge.
Benefits of off-chain processing and zkSNARKs are scalability and privacy.
Vitalik gave a very excited presentation showing what he sees being in store for the next 2–4 years for Ethereum. He laid out a case for the long term scalability solution for Ethereum being sharding (“Quadratic Sharding”) — where there are N Ethereum shards, each essentially having its own chain. Transactions can be executed between shards, and Ether and other digital assets can be transfered between shards.
Vitalik also discussed the “Trilemma” of blockchain systems
in that it’s generally considered true that blockchain systems can only achieved 2 of the following 3 properties:
Scalability
Decentralization
Jump to Summary
