Lately I've been thinking about cross-chain bridges again. To be honest, no matter how lucrative the profits are, it all depends on whether the "bridge" will send you flying with a shovel. Many bridges are hyped up with fancy features, but I prefer to focus on multi-signature: who signs, how many keys, whether there's a time lock. Without these, I just treat it as a warehouse that can open a back door at any time. Oracles are the same—once the quotes/status are skewed, the other side of the bridge will also misjudge, and it's too late to react when something goes wrong.

Then there's that old cliché of "waiting for confirmation"... I used to think it was slow, but after losing money once, I became honest: when it comes to cross-chain steps, it's better to be a bit slow, wait for a few more confirmations, at least leave some time to retreat. Recently, people have been complaining about validator income, MEV, unfair ordering, and so on. I haven't really kept up either; honestly, just hearing the word "ordering" gives me a headache... But it seems like within the same chain, you can be caught in the middle, let alone crossing to another chain. My approach is still the gardener's method: use the minimum amount for the bridge, avoid crossing if possible, and if you must cross, do it in batches, wait for confirmations, and don't operate before bed.