Lately, on-chain data keeps getting “stuck” for a moment. Others think it’s their bad network or their wallet acting up—but in reality, a lot of the time it’s the indexer/subgraph catching up on blocks, recomputing, or the RPC getting rate-limited. The response comes late by half a beat, or sometimes even goes blank entirely. Especially when a new L1/L2 comes out and uses incentives to pull TVL, the call volume spikes; over on the node side, they first queue up all the small users’ requests… You watch the price/depth change like it’s teleporting, and by the time you finally place the order, the slippage is maxed out. To put it bluntly, it’s like you’re being treated as a liquidity donor—except the “donation” is coming from your execution.

Anyway, I’d rather just wait an extra few seconds to confirm: compare the same piece of data across two RPCs, don’t trust route estimates too blindly, and if it’s truly urgent, first factor in the impact cost—otherwise you’ll just get “educated” by the whole cycle of digging up, raising, and selling. For now, that’s how it is.