I was a bit scared just now: I saw on-chain that a transaction "was not successful," and I almost wanted to retry it. Luckily, I held back and switched to a different RPC, then checked again—turns out the same transaction has already been included in a block, it just hasn't synchronized to my usual node/index yet... Basically, what you think you're seeing as the "truth on the chain" is actually the perspective and latency of a certain service.

Recently, that mainstream public chain has been undergoing upgrades and maintenance, and everyone in the group is guessing whether the project will migrate. I think there's no need to jump to conclusions just yet; node restarts, RPC congestion, and index lag are quite common before and after upgrades. The data you see might be a bit behind. My quick fix: before and after critical operations, cross-verify with two or three different entry points (different RPCs, different browsers). If possible, wait a little longer—it's better to be slow than to accidentally fall into the trap of "repeated authorization/repeated transactions."