[kang]

> A smart contract is simply an agreement between parties, written in codes and enforced by a third-party.

No, it is an agreement between parties enforced by cryptography, and no third party required.

> In the case of Ethereum, it is enforced by cryptography. No it is enforced by payment to the network, just like aws, but in a worst way, because code is executed wastefully, while still remaining at a central location.

What you are not understanding is that a smart contract is a protocol amongst people. You don't code it anywhere in the blockchain directly. It a a protocol that explains how, if each party runs code independently, assuming everyone as a criminal, still achieves its goal. For eg, multisig is a smart contract, which is a protocol where participants decide what code they'll run (signing) and the wallet layer that makes them meet is a separate layer altogether, not in the blockchain. Sometimes this layer may not even be required, for example in money-laundering or any of the 'hundis'.

[ncodes]

> What you are not understanding is that a smart contract is a protocol amongst people. You don't code it anywhere in the blockchain directly. It a a protocol that explains how, if each party runs code independently, assuming everyone as a criminal, still achieves its goal.

Okay, I see your point. This kind of smart contract you describe definitely needs strong cryptography and a blockchain would not be needed. I guess the issue is still about what a smart contract truly is and how it should work. I personally prefer the term self-executing contract as the "smart" seems too sensitive. Regardless of the language or definitions, I am interested in a platform were multisig applications can be executed safely and efficiently.