Perp protocol is going with xDAI chain. I don’t know if that is a possible scaling solution for YFI.
Yes you´re right, I was a bit unprecise. By “port over” I actually meant copying it onto L2. Scaling solutions like Matic are fully EVM compatible so they only need a redeployment of the contracts. Which is dirt cheap there, so it´s a no-brainer.
Concerning xDai I think they are not 100% EVM compatible, but I don´t remember clearly. Maybe that process needs some more work.
The communication between different L2 solutions will be an interesting topic for the future. Perhaps the interoperability blockchains like Polkadot, Kusama and Cosmos bring a meaningfull solution. Or possibly there could be some transaction coordination so that the information exchange can be batched and done only once every few hours or so over L1? I´m still in the process to learn more about how L2 actually works, so if someone knows more, please chime in.
@ejbaraza Yeah xDai and Matic seem to be the two major L2 solutions currently, with slightly different tech (plasma vs. sidechain). xDai has a bridging process as well, so like with Matic it´s kind of its own little universe. Unless there can be some fast and cheap communication with L1 (which I doubt, why would you need a bridge then), it´s not really a scaling solution but rather a “branch-out” solution. I´m definitely excited about these branch-outs though. Everyone here should really try out L2 solutions. You wont like to get back to L1 afterwards. 
According to Perp, the xDAI sidechain is a temporary solution. If/when some better L2 hits mainnet they said they are going to take a snapshot and port everything over to said L2 solution. So it might be a good short-term solution to gas costs and "bridge the gap to ETH 2 since xDAI is in the process of going PoS.
2 Likes
Do you have a good source to read about Matic? Especially the issue of being fully interoperable with L1.
If you’re saying that we literally only need to redeploy contracts on Matic L2 and everything will work fine and we’ll pay low low gas, why wouldn’t more people be using it already? Honest question, not trying to be snarky.
1 Like
At this time, incorporating Layer 2 is impractical for Yearn’s goals and how it integrates with the rest of the ecosystem, as some have metioned above. The only place a L2 system might be useful is in designing the governance system that would connect to the Vault (eventually). That governance system is out of scope of this design proposal, so we can safely ignore L2 design considerations in this proposal!
Feel free to discuss L2 in another proposal or discussion thread besides this one.
2 Likes
I feel like this may have been mentioned before @fubuloubu, but will v2 yVaults be able to accept deposits from more than one token? Or is this something that can instead be built into a strategy?
For instance, will we still need a separate aLINK and LINK yVault? Or will the strategy or vault parameters be able to be adjusted to accept any reasonable token for the strategy?
In that sense, yes, gas tokens are just rent-seeking. It’s like booking all the tables at a restaurant to resell without ever intending to eat there.
I agree that L2 solutions are needed (L1 congestion will be a problem for at least a couple years) but a particular silo-ed L2 solution limits our ability to seek out the best strategies. I’m not sure.
It goes back to simplicity.
We want people to be able to come to one front-end, and manage their assets there.
In this sense, I think (and this is just one perspective) that the simplicity of the user flow of deposit x, you get back yx, withdraw x + returns is strong.
On the other hand… At the same time, any Aave-token vaults (aLINK) or wrapped token vaults (wETH) or other pool token/derivative token (yCRV, y3) requires an intermediate step on the part of the end user before deposit.
Acquiring the pooled and/or wrapped tokens for some of the vaults is a giant pain in the ass. Simply investing in the yUSD vault is an intimidating process for a DeFi-naive user, and still not a perfectly simple one for someone well-versed in DeFi. Depending on liquidity and gas costs and how many different protocols you’re familiar with and following, you could enter in a variety of ways: deposit to Curve (but be sure not to stake in the gauge!), and then deposit in Yearn; use a separate front-end like Zapper (but pay the gas costs for having a smart contract do it for you); swap on an AMM like Uniswap if you can find liquidity (or Snowswap, but they don’t keep pace with the newest, latest, and greatest vaults…) directly into yUSD tokens.
It’s a tough tradeoff. Because ultimately, some strategies are better suited/only available directly and not through delegation to certain assets. And all the stuff I was describing above…they can be gas-expensive, not just user-time/experience expensive.
Ultimately, yLINK deposits were suspended because having an additional layer to wrap LINK on behalf of the end user cost a lot in terms of gas. wETH/ETH is a lot cheaper, but every time we have multiple options for deposits of an asset we introduce a lot of complexity.
I think there’s a sweet spot, and we still need to find it.
No, they are single asset. The additional protections in the new design allow them to basically go long on their underlying through multiple Strategies, such as using aLINK for LINK
2 Likes
What is the future of the USDT vault?
I haven’t been able to find out, so hope someone will share some info here.
If not where can I ask instead?