[rfd4sgmk8u]

Disagree. LN providers will exist, just as exchanges exist today. The most likely candidate is the payment card networks. While I dislike this future, they probably offer the best solution for consumers interacting with layer 2, and its always on requirements. LN isn't just a toy, its a new way of transferring value peer-to-peer. Payment/state channels are the best candidate to scale this tech. Anything else -- sidechains sharding or block size increases is just kicking the can. Not every transaction must be recorded on a blockchain forever. Layer 2 is how we will achieve this, while retaining the massive value that Layer 1 provides. I also find technologies like cut-through transactions in mimblewimble compelling for scaling without recording everything.

You must be joking re eth scaling. ERC20 tx's presently cost $20USD https://etherscan.io/gastracker A low priority btc tx currently costs 57 sat $4.40 https://mempool.space/

Currently 5x the price to transfer a fake bitcoin token on eth to a BTC L1 tx, whereas transferring a real bitcoin on lightning might cost a few satoshi.

I think some elements of eth are interesting, but are unlikely to actually achieve whatever it is they want to achieve. Proof of Stake isn't secure. Lets see if they are able to retain consensus with EIP-1559. I suspect it will cause another schism.

[CryptoPunk]

>>You must be joking re eth scaling. ERC20 tx's presently cost $20USD https://etherscan.io/gastracker A low priority btc tx currently costs 57 sat $4.40 https://mempool.space/

These are not using the recently launched scaling technologies. For example zk-Rollups, which are now operational on Ethereum mainnet, can process 2500-3000 ERC20 transactions a second on-chain. Once these gain a critical mass of users, Bitcoin, in the form of WBTC, can be widely used in payments by way of Ethereum's zk-Rollups.

The LN's limitations are too many, and it has thus far failed to gain traction for that reason. Zk-Rollups have none of the limitations of off-chain payment-channel networks.

[rfd4sgmk8u]

zkrollups, like all rollup technologies just kick the can down the road. Not every transaction needs to live on a blockchain! This is the fundamental point that folks miss. LN is about true 'off-chain' in that a blockchain is not needed to provide a probabilistic model to funds. Do you see how this is similar to MW cut-through, in that a probabilistic model is used to store funds on chain? Imagine all the ZKrollup solutions and throw away the sidechain.

Rollup is just another name for sidechains. I agree, the sidechain model works today. The BTC Pegged liquid network also does zero knowledge (confidential transactions) exchanges of BTC between exchanges. But ultimately it is still kicking the can down to another blockchain. Rollups are not 'off-chain', They are on another chain. We need to think beyond this for true scaling.

[CryptoPunk]

LN not having data on-chain imposes severe constraints that make LN inherently more centralized and less secure than zkRollups.

If you're really set on payment/state channel networks like the Lightning Network (Bitcoin), Raiden (Ethereum) or Connext (Ethereum), you can put them, ON A ZK-ROLLUP!

You get a 100X scalability boost, without increasing the block size, which makes payment/state channel opens/closes a couple orders magnitude lower, which massively increases a payment channel user's options and improves the functionality of the payment channel network. ZkRollups also reduce the cost of on-chain payments which lets a user switch to on-chain for an application that the zkRollup-based PCN is ill-suited.

All-in-all, reducing the cost of on-chain transactions, as zkRollups do, have significant advantages, so there is absolutely no reason to not adopt them.

And Liquid is a centralized sidechain dependent for its security on trusted third parties, and completely incomparable to zkRollups, which have 100% of the security of the mainnet and are only reliant on 51% of the hashrate being honest for security.