Recently, some people have been frightened by a bunch of buzzwords like “data availability / ordering / finality.” In fact, you only need to focus on one main thread: when you send out that transaction, is it really about whether others can see the data, how ordering is determined (who comes first and who comes later), and whether it will ultimately “go wrong.” What cross-chain / routing fears most is not being slow—it’s when you think things are already stable, but the ordering around that chain’s maintenance or upgrades gets shaken, finality drags on longer, and over on the bridge things start to get stuck, to retry, or even to execute repeatedly. Now everyone is guessing whether the ecosystem will migrate after a certain mainstream public chain upgrade; I think it’s better not to rush to declare a migration. First, see which part the upgrade makes more complicated: can the data be seen, is the ordering correct, and does it settle reliably. One more reminder: I treat “simplicity” as a trap—statements like “it finalizes quickly” or “it basically never rolls back” are best taken as something you listen to, not something you treat as a design premise.