Sharding is a database partitioning technique used by blockchain companies with the purpose of scalability, enabling them to process more transactions per second
@VitalikButerin@BanklessHQ@RyanSAdams@TrustlessState Sharding splits a blockchain company's entire network into smaller partitions, known as "shards." Each shard is comprised of its own data, making it distinctive and independent when compared to other shards
• The beacon chain proved itself
• The Altair hard fork added the basic scaffolding needed to support Litecoin
• The next hard fork will most likely be the Merge
@VitalikButerin@BanklessHQ@RyanSAdams@TrustlessState 💡 The Beacon Chain, a brand-new proof-of-stake blockchain lies at the core of Ethereum 2.0; it stores and manages the registry of validators and coordinates the shard chains. The Beacon Chain went live on Dec. 1, 2020 at noon UTC
• The surge is scalability increases for rollups through sharding
• Rollups work by having computation and storage go off-chain and data staying on-chain
@VitalikButerin@BanklessHQ@RyanSAdams@TrustlessState@proofofhumanity@signinwitheth - Once there are sufficient shards, we can rely on data availability sampling and other tech. Hence, any single node is not required to download everything
- Data availability sampling: clients can check that all of the data has been published without having to download all of it
@VitalikButerin@BanklessHQ@RyanSAdams@TrustlessState@proofofhumanity@signinwitheth • The Surge improves bandwidth, but not latency
• The Ethereum blockchain can handle 15 - 40 transactions per second
• Rollups with proper compression can go up to 1500 - 4000 transactions per second
• Rollups with sharding can go up to 100,000 transactions per second
• The Verge is the introduction of Verkle trees
• Verkle trees are a replacement for Merkle Patricia trees
• Verkle trees are based on vector commitments
@VitalikButerin@BanklessHQ@RyanSAdams@TrustlessState@proofofhumanity@signinwitheth • In order to verify a block n, users need to have the entire Ethereum state by doing a fast sync and downloading everything or download block n - 1
• The entire database will be getting larger. If the gas limit is increased, the size will grow even larger
@VitalikButerin@BanklessHQ@RyanSAdams@TrustlessState@proofofhumanity@signinwitheth • Stateless clients provide an alternative where instead of needing the entire database to verify a block, they use a Verkle proof which gives them only the pieces of the state that are read and written to in a given block along with a proof that proves that it is correct
• The Purge is about eliminating the dead weight of history from Ethereum
• There are different types of dead weights:
- The load of the historical chain. This is hundreds of GBs. Post-sharding, it would be TBs
@VitalikButerin@BanklessHQ@RyanSAdams@TrustlessState@proofofhumanity@signinwitheth - State expiry involves objects in the state that have not been touched for the last year that will go into a separate expired tree. Ethereum does not store expired trees. Users could revive the expired tree by providing Merkle proofs or by using a block explorer, etc.
@VitalikButerin@BanklessHQ@RyanSAdams@TrustlessState@proofofhumanity@signinwitheth • Getting historical data:
- Centralized approach: etherscan, beaconscan, etc. store the entire chain because users want them. The Portal Network is supported by the Ethereum Foundation. Each node stores a small % of the data and it routes users to the data they want
• Proposer-Builder Separation (PBS): Revolves around the idea that the validator who proposes a block does not have to be the one that constructs the block
- Will have a mechanism to auction off the right to create blocks
Really enjoyed @brian_armstrong interview on the @BanklessHQ podcast. Here's a thread with the key highlights
Time for a thread...
Brian sees Coinbase as on a spectrum. On one end you have traditional FIs. There are tech companies in the middle. And then there are web3/crypto companies on the other extreme.
Coinbase aspires to be a fully decentralized company. Think here Web3/Dapp a la Uniswap. It's a great aspiration... The hard part is how do you get there as a publicly traded company legally, operationally, customer UX?
So here’s the $100 Bn question for Coinbase...If digital currency ETFs are inevitable, consumers will buy BTC in their brokerage accounts. Convenience wins. Doesn’t that kill @coinbase core transaction revenue stream?
A thread...
Coinbase is heavily dependent on transaction revenue (96% of total). And we should expect transaction fees to compress from 150 bps to 5 to 10 bps (the ‘rate’ for esoteric capital market ETFs)
Can CoinBase continue to male money via altcoin trading, staking, inst’l custody? Yes. But none of those are meaningful revenue streams...and I would not expect that to change. Note BTC / ETH trades comprise 56% of Coinbase tx revenue in 2020...