[dinobones]

Thank-you for the helpful links. Can you share some resources to learn about data availability sampling?

Also, have folks invented a cheap/fast way of going from L2 <-> L2 without having to do an L1 tx?

I fear that L2s may never be adopted due to network segmentation, but if it's possible for all L2s to interchange with each other cheaply, then it's just as good as L1 IMO.

[everfree]

This post from the ethresearch forum goes over data availability sampling (DAS) in detail.

https://ethresear.ch/t/from-4844-to-danksharding-a-path-to-s...

To transfer assets from L2 to L2, of course the naive implementation is to use a centralized intermediary, of which there are currently many that are reasonably priced. There are ways to go between zk-L2s without any central broker in theory; I’m not sure whether that’s also true of optimistic-L2s.

[chrisco255]

> have folks invented a cheap/fast way of going from L2 <-> L2 without having to do an L1 tx?

There are bridge providers like Connext (https://www.connext.network/), Hop (https://portal.arbitrum.io/projects/bridges-and-on-ramps?pro...), LayerZero (https://layerzero.network/) etc that provide liquidity across L2s to make it simple and cheap for common assets like USDC, ETH, etc.

Attempts to do this trustlessly without relying on a liquidity provider do exist, but they're not mature enough to mention yet. They usually rely on zk proofs to validate that an asset was bridged from one chain to another.

L2s are presently already supporting more activity than L1, with 4 L2s regularly doing more TPS than L1. Agreed that fragmentation is a concern, but I think we'll get there soon where the UX is abstracted away for users and the assets flow cheaply.

[TimJRobinson]

I recently wrote a post covering all the ways L2s will intoperate and share liquidity with each other: https://paragraph.xyz/@blueyard/how-ethereum-can-solve-l2-li...