Doesn't this just make on-boarding users even more convoluted? You'll have to probably get a large amount of Ether and then 'commit' it to another L2 to receive whatever their backed token equivalent is. Then you'll use this token to pay for things which will have no network effect, apps, or merchants in the beginning. Basically all adoption goes back to square one while providing a confusing and worse user-experience for everyone. How are wallets supposed to explain this shit to users?
You're right, it sucks in the short term. The way I look at ethereum/defi is that we're interacting with the protocol directly right now. Imagine if you could only use your bank by working directly with the API, and then some people hacked together crude tools to make the API more tolerable.
Over time this will all be abstracted away and typical users won't have to worry how it's all working underneath (unless they really want to).
Even if you can withdraw straight to L2s, the main problem I have with this scaling solution is it is too convoluted. I have no faith in being able to explain this to people who aren't already crypto enthusiasts. It seems much easier to explain how to use a single scalable L1 than to explain a cluster of ever growing L2s and sidechains. I know there is a lot of work in the ecosystem but really this has to be masked hard with a fantastic web wallet UI if it has any hopes of widescale and longterm adoption imo. This is why I have more hope currently for Solana than Ethereum.
It's an incredibly complicated problem that is currently being worked on by ETH developers. However eventually this will be seemless for an average user. You can be using Uniswap for example on an L2 like arbitrum or zksync and have no idea your ETH is even there. Also ramps are being built to directly deposit into L2. The gas fees from bridging in and out of L2 will be solved when the majority of the traffic is happening on L2, the ETH L1 blockchain will only be used as a settlement layer for L2 traffic and much fewer transactions. Sharding which is coming will also bring down fees a ton for all settlement transactions in ETH L1.
May I ask what is probably a stupid question? What about Phantom.app as a wallet for a noob? I am leaning towards ETH, but all of the wallets looks exceedingly complicated...except for this one.
Phantom wallet is great! It's currently primary a solana wallet, but I believe that they are in beta for Ethereum support, but if you are curious to try out web3 stuff Solana is great for it. I strongly prefer the UI over this wallet compared to Metamask for instance, and think it is a very user friendly wallet.
Most of them use ETH as their currency, but even if they have their own token once exchanges have direct withdrawls to them, and they have most apps, they'll feel the same as using any other L1.
We're still very much in the early days of this though, having to move between chains manually. There are bridges that allow executing contracts on other networks, and eventually apps will abstract the different layers away from users using technology like this.
So once again, Ethereum users relying on centralized exchanges to bring the gas fees down (when the steps involved confuses users and the fees are still high).
Given that being the case, the general decentralization argument would be pointless. For the many users who need to be able to pay for their groceries quickly, you might as well use Solana then, since everyone needs to eat.
The whole point of Uniswap is to move away from these centralized exchanges. Perhaps using that has gotten more complicated to use or the swaps have also gotten too expensive even if it is on an L2; making that useless as well?
Arbitrum doesn't require you to use a centralized exchange, it's just easier to onboard with "fiat -> centralized exchange -> arbitrum" than "fiat -> centralized exchange -> ethereum -> arbitrum".
