Problem: OOKI does not have a price feed, so there isn’t utility around it: Ooki token can’t be traded, lent, or borrowed against.
Solution: OOKI uses API3 to get an OOKI price feed. This price feed is relatively easy to spin up via API3, so no significant time away from devs.
We have been waiting for years on ooki price feed from chainlink. So I agree moving forward for another oracle. And this should be done asap!
I fully agree with you, we can’t wait for chainlink any longer.
After careful research, I have some questions before committing either way for integration. I would like to preface that the idea is great and could really be useful for the protocol in the long run. However, there are some questions I have regarding price feed security, decentralization, and potential future price increases for feeds.
Security: My main question regarding this is if the main publisher for a price feed were to go offline, is there some sort of backup to maintain the feed? The protocol makes a major trust assumption on the feed to be actively updated 24/7 and the single provider solution offered for feeds on API3 cause some worry on my end of upkeep. Likewise, are there any built-in price bands to restrict the acceptable published prices on the feeds? Another concern is the provider accidentally pushes wildly inaccurate data causing liquidations. Although there is insurance provided to cover this as previously stated, which is a very desirable feature, it is still a nuisance and reduces trust in the protocol, and distributing the claims is likely cumbersome. It is not unprecedented for reliable firms to report wildly inaccurate data (for example Deribit’s index rate blunder years ago). The price band setup aids in restricting how far the price can be reported from the previous one which nullifies most inaccurate recordings due to bugs.
Decentralization: Reading through the docs page I noticed a section about a decentralized oracle that seemed far more intriguing than a single provider. The decentralized oracle actually addresses most of my concerns and I was wondering when/if that will be offered and the cost differences associated? Likewise, how many nodes are sampled for gathering the data? One of the main goals for the protocol should be to ensure it remains decentralized and that includes the protocols that act as dependencies. Having a decentralized oracle feels like a must in my opinion
Monthly costs: Is there some sort of breakdown for what dictates how much is charged monthly? One thing I want to see is what is the cost breakdown so it is easy to predict how much things may cost on a chain if gas costs were to spike and to factor in that risk for a feed. The main thing is canceling a feed while loans are open is quite difficult and would require a forced liquidation event which is undesirable for UX so to avoid that, only select pairs would be focused on those that generate more fees than the cost for the feed. This means the future feed costs need to be thought of measured as a potential situation.
It would be greatly appreciated if some of these concerns could be addressed and we can go forwards from there. Thanks!
Edit: Serves as a response to OOKI to obtain any Price Feeds via API3 - #10 by API3OOKs as well since the standards for integration for one feed or many is the same.
Why not use dextool or poocoin?