4 ms·

> Money: the first and still most important app Money is also the thing that will continue to be problematic for Ethereum going forward. Two technical problems

> Money: the first and still most important app Money is also the thing that will continue to be problematic for Ethereum going forward. Two technical problems drive this: 1. Scalability issues related to a very large block chain size and the associated difficulty in bringing up and running a full node. 2. The transition from proof-of-work to proof-of-stake is driving the convergence of validators and exchanges. (1) drives individuals increasingly into the arms of gatekeepers to tell them what's "on the block chain" rather than directly participating on the network. (2) degrades what little censorship resistance Ethereum may have once had.

x-complexity 4y ago
> 1. Scalability issues related to a very large block chain size and the associated difficulty in bringing up and running a full node. This is a problem of 2 parts: Transaction scalability & Storage scalability. Transaction scalability's currently being resolved with various optimistic & zk-based rollup solutions, with the latter rolling out next year (zkSync 2.0, Scroll, & Polygon Zero & Miden) and the former already being improved upon & decentralized (Optimism & Arbitrum). There are still guardrails in place, as they're still being developed/improved on. https://l2beat.com/scaling/risk https://l2beat.com/scaling/risk Storage scalability's a different issue that's also being worked on via data availability (DA) sampling, whereby the use of erasure codes & random sampling allows for data to be split into chunks & distributed amongst the network. With this system, every node doesn't need to hold onto all of the data: The whole network only needs to provide at least 51-99% (depending on the erasure code used) of the chunks to reconstruct the block, verify, & republish the block for everyone else. By doing so, the average case for a node is reduced from "store all data" to "store 1/n of all data", with N being the number of data shards for the entire network. Another way that's being worked on is the creation of Validium rollups, wherein the data is stored off-chain but it can still be proven via proofs published on-chain. It's not as secure as a fully on-chain rollup, but in return, the rollup consumes significantly less gas. https://twitter.com/VitalikButerin/status/1588669782471368704 https://twitter.com/VitalikButerin/status/158866978247136870... https://pbs.twimg.com/media/FgwVhUjaAAEx_Bb?format=jpg&name=large https://pbs.twimg.com/media/FgwVhUjaAAEx_Bb?format=jpg&name=...
x-complexity 4y ago
> 2. The transition from proof-of-work to proof-of-stake is driving the convergence of validators and exchanges. ...I have no idea what's being discussed here: Can you explain further into detail? My best guess on this is that you're saying that exchanges, as a result of being the places where non-enthusiasts will park their their money at for safe investments into this space, will consequently be the place where we'll see more validators come online at, resulting in a concentration of network power accruing at the exchanges. This has been a known problem with its own solutions, with Rocketpool being the leading case for decentralized staking. It's not perfect, but it allows for people to stake into a pool, & for validators to get the ETH needed to stake a validator node in a non-centralized manner, with improvements & lower validator requirements already being worked on. Stakewise is also going in this direction with DVT for V3. https://rocketpool.net/ https://rocketpool.net/ https://twitter.com/stakewise_io/status/1575487080008130562 https://twitter.com/stakewise_io/status/1575487080008130562

x-complexity 4y ago

> 1. Scalability issues related to a very large block chain size and the associated difficulty in bringing up and running a full node. This is a problem of 2 parts: Transaction scalability & Storage scalability. Transaction scalability's currently being resolved with various optimistic & zk-based rollup solutions, with the latter rolling out next year (zkSync 2.0, Scroll, & Polygon Zero & Miden) and the former already being improved upon & decentralized (Optimism & Arbitrum). There are still guardrails in place, as they're still being developed/improved on. https://l2beat.com/scaling/risk https://l2beat.com/scaling/risk Storage scalability's a different issue that's also being worked on via data availability (DA) sampling, whereby the use of erasure codes & random sampling allows for data to be split into chunks & distributed amongst the network. With this system, every node doesn't need to hold onto all of the data: The whole network only needs to provide at least 51-99% (depending on the erasure code used) of the chunks to reconstruct the block, verify, & republish the block for everyone else. By doing so, the average case for a node is reduced from "store all data" to "store 1/n of all data", with N being the number of data shards for the entire network. Another way that's being worked on is the creation of Validium rollups, wherein the data is stored off-chain but it can still be proven via proofs published on-chain. It's not as secure as a fully on-chain rollup, but in return, the rollup consumes significantly less gas. https://twitter.com/VitalikButerin/status/1588669782471368704 https://twitter.com/VitalikButerin/status/158866978247136870... https://pbs.twimg.com/media/FgwVhUjaAAEx_Bb?format=jpg&name=large https://pbs.twimg.com/media/FgwVhUjaAAEx_Bb?format=jpg&name=...

x-complexity 4y ago

> 2. The transition from proof-of-work to proof-of-stake is driving the convergence of validators and exchanges. ...I have no idea what's being discussed here: Can you explain further into detail? My best guess on this is that you're saying that exchanges, as a result of being the places where non-enthusiasts will park their their money at for safe investments into this space, will consequently be the place where we'll see more validators come online at, resulting in a concentration of network power accruing at the exchanges. This has been a known problem with its own solutions, with Rocketpool being the leading case for decentralized staking. It's not perfect, but it allows for people to stake into a pool, & for validators to get the ETH needed to stake a validator node in a non-centralized manner, with improvements & lower validator requirements already being worked on. Stakewise is also going in this direction with DVT for V3. https://rocketpool.net/ https://rocketpool.net/ https://twitter.com/stakewise_io/status/1575487080008130562 https://twitter.com/stakewise_io/status/1575487080008130562