Possibly the worst decision I made back then when I was buying stupid easy domains... Sometimes I wonder if I'm doing the same now... I have made 5 trades in the last week and 4 were bangers but I almost avoided them because of crazy fees. Maybe fees are a signal? #ETH#Crypto
Ok, lots of discourse on L2s/L3s so here's an overview for those of you that are confused.
TLDR: L2s are great but flawed right now, L3s are highly subjective, and we should stop being hypocrites and calling Solana centralized.
L2 Overview:
First, we need to understand what L2s are. They scale base layer blockchains (L1s) by executing transactions on the L2, batching the tx data, and compressing the data.
The compressed data is sent to the L1, and is used in either the fraud proof (op rollup) or validity proof (zk rollup). The proving process is referred to as "settlement".
L2 Risks:
Most rollups currently have a single, centralized sequencer. The sequencer has the ability to censor transactions, although tx force inclusion is available on most rollups.
The proving systems on L2s are also currently in the "training wheel" phase. Fraud proofs are currently permissioned (Arbitrum) or not present (OP stack). ZK proofs are all permissioned. This reduces the trust assumptions, or security threshold of rollups.
Additionally, the bridge contract that holds all the rollup's bridged ETH is currently upgradeable. This means that the ETH in the bridge can be stolen (if either the DAO or multisig approves of it).
L3s:
These are basically just L2s, but they use a rollup as the base layer. The caveat however, is that the tx data that's compressed and sent from the L3 to the L2 does not get further compressed when the L2 posts data to the L1. So in that regard, there's no significant scaling/cost benefit.
An L3 is not "real" in the sense that it doesn't further scale the L1. It's really just an L2 ontop of an L2.
L3 Risks:
L3s have the same centralized sequencer and bridge upgradeability concerns as L2s. However, the base layer for L3s also have the same concerns. Unlike an L2 where the base layer (L1) is decentralized and censorship resistance, an L3 relies on base layer that has several security concerns currently.
What this all means:
L2s have a lot of maturing to do before we can really consider them secure and at the level of "inheriting security from Ethereum". There are legitimate centralization and upgradeability concerns. These can and most likely will be fixed in the future.
I do think it's hypocritical to call Solana centralized while L2s have these risks. Solana has a decently large validator set and cannot have funds stolen by a simple bridge upgrade. Both Solana and various L2s have suffered downtime in the past. There's work to be done on both fronts, and I argue that both are valid ways of approaching blockchain scaling.
The value add of L3s are more subjective. There are some marginal cost benefits, alignment benefits, and some room for experimentation. However, I can see why many people are frustrated with the L3 narrative - we haven't reached maturity with L2s and are now pitching a whole new layer with even weaker security. Time will tell if the people care about any of this.
Finally getting around to checking out some of these blockchain based social media networks.
I'm votetrev.op.cast on https://t.co/pyFCEI5zSS built by @wieldlabs exploring Farcaster, join me to earn FarPoints! https://t.co/ZZssmTcgoX
@gxdia@gxdia please stop showing these pictures of our food. This is one of the last shrimp puddles we have and if anyone with a brain CSI's into the sand pic they can tell we are on the west side. STOP X'ing!
Anyone using the https://t.co/VEqf4wkNVp db with AI to do anything interesting? Been having some fun with it and cold email and would love to hear anything else interesting being done :) #coldemail#apollo#ai#deez
@TheTinyThug@OneOnesNFT Nice move, I had a good feeling about this project and minted 2. How did you end up minting 3? My first guess was you had two addresses whitelisted, but then I assume you would have minted 4 and not 3.