Etherium is a first cut at a hard problem. The "smart contracts" problem is interesting, but after the DAO debacle, it's clear that smart contracts shouldn't be executable programs. It's too hard to predict their behavior, and that's very bad when they're connected directly to money. Some human-readable declarative notation is needed. I've suggested decision tables. Dealers need enough expressive power to handle all the usual financial instruments, but don't need or want Turing completeness.
Another advantage of a declarative notation is that compute time tends to be bounded by a fairly low bound. That would put a lid on problems like this new one, where the little programs are chewing up too much time.
We may see some smart contracts system with a better notation in the derivatives business. It doesn't need to be a currency, just a blockchain with enough identified independent players to prevent tampering. The need for "mining" comes from the desire for anonymity; it's not inherent in smart contracts.
Another advantage of a declarative notation is that compute time tends to be bounded by a fairly low bound. That would put a lid on problems like this new one, where the little programs are chewing up too much time.
We may see some smart contracts system with a better notation in the derivatives business. It doesn't need to be a currency, just a blockchain with enough identified independent players to prevent tampering. The need for "mining" comes from the desire for anonymity; it's not inherent in smart contracts.