Different DEX design questions

[ducksquaddd]

I have a DEX that doesnt use the standard pricing model most DEX's use and we actually require an oracle price from PYTH to provide an accurate swap rate.

So my question is whats the best way to go about doing that within this kyberswap integration. Am I fine to simply query a pyth contract for the price of both assets on every request if we have to go the onchain route?

[NgoKimPhu]

If your DEX can handle price deviation up to 2% you can directly use Pyth's Push Feeds to query directly onchain, so in the integration you also just track this through rpc calls to the pyth contract and saves this in the pool state.
If you want to go the Pull Feeds route, then you can take a look at BrownFi V2's integration where the offchain Pyth data is fetched and tracked in the pool state and later used onchain to push to pyth oracle before use

[99adarsh]

If your DEX can handle price deviation up to 2% you can directly use Pyth's Push Feeds to query directly onchain, so in the integration you also just track this through rpc calls to the pyth contract and saves this in the pool state. If you want to go the Pull Feeds route, then you can take a look at BrownFi V2's integration where the offchain Pyth data is fetched and tracked in the pool state and later used onchain to push to pyth oracle before use

Hey @NgoKimPhu, our DEX's architecture is similar to BrownFi. Instead of integrating through the AMM simulation approach, where we would have to write the entire AMM calculation logic for simulation, can we integrate using the RFQ approach? We will maintain a server that returns the amountOut along with transaction data as QuoteResult, and also returns the prices that can be used in simulation to calculate the amountOut.

[NgoKimPhu]

we only use RFQ if the data is completely offchain and can't possibly be fetched thru onchain methods