To understand, why The DAO failure is NOT smart contract failure, one needs to understand a few points.
What happened with The DAO was just a piece of code which was not written as it was intended to work. The real problem right now is the interest of Ethereum Foundation to save The DAO and hence lobbying for a hard fork. The DAO was supposed to fund further development of Ethereum and hence the interest is. Ethereum Foundation did not speak about a hard fork when ShapeShift was hacked. This clearly shows their conflict of interest and if the hard fork is successful, it would prove the centralized point of failure of Ethereum Blockchain.
On the contrary, after the Mt. Gox debacle, there were many people crying on bitcointalk to hard fork the bitcoin blockchain to save their coins. But, no one could centrally influence the diverse bitcoin mining community. In reality, no hard fork ever took place in bitcoin blockchain to revert any theft, heist or whatever. This only proves that the bitcoin blockchain is immutable.
Now, a smart contract needs to be written on immutable blockchain. If any party has the authority to reverse a smart contract, it is not that smart anymore. Hence, I believe, irrespective of the future of Ethereum, smart contract do have a future on the immutable bitcoin blockchain through secondary layers like Rootstock or Counterparty.
Thanks for the comment:
There are lots of doubts about the code to be categorized malicious at all:
some point fingers to the interpreter itself and others put forward a more philosophical question about who is authorized to classify a code as being corrupted and the hacker himself has claimed to be fully authorized and legitimate to suck 200 M$ out of DAO and ...
DAO failure is not just the failure of a piece of code it is exactly the failure of the 'signing' process that handed the power to that malicious code to be run on behalf of the people. This is the problem: Turing complete code is full of tricks and traps how can we be 'smart' enough to verify a contract coded in a Turing machine language? Can we ever be such a smart person? I doubt it.
And the hard fork Ethereum Foundation is determined about, is not a good practice but is legitimate. I mean this is what hard or soft forks are all about. If they can gather enough votes, a consensus, they should manage for the fork but it can't enhance the morale of the smart contract enthusiasts as not every failure will be such a huge dramatic one and worth a fork.
