4 ms·

For example they had that bug last year where somebody was able to steal 50 million dollars equivalent ether from the network. IIRC that caused one of the hard

For example they had that bug last year where somebody was able to steal 50 million dollars equivalent ether from the network. IIRC that caused one of the hard forks and is the reason there are now both eth and etc. That was a bug in the DAO protocol. I'm not aware of any bugs in the btc protocol to allow hackers to steal money in their entire 8 years existence

verroq 9y ago
It wasn't a bug with the protocol though. It was a bug with the coding of the contract. However there has been a few forks to mitigate DDoS attacks.
- gnaritas 9y ago
  No, it was worse; it was a bug in the community that proved people aren't really willing to abide by smart contracts when they feel they lost. That incident split the community in half with one side willing to tear down the entire thing just to make sure the thief didn't get away with it. And now there are 2 Ethereums.
  - codingmyway 9y ago
    Ironically that might be a reason for the later increase in the price of ETH compared to ETC as corporate types like the idea of being able to 'roll back' the chain in the event of fraud or a serious bug and don't appreciate that the DAO fork was a one-off event where the majority of the stakeholders were in favor due to their foolish purchase of DAO tokens.
tscs37 9y ago
DAO wasn't a protocol anymore than a Bitcoin Client is part of the protocol. It was an application that failed.
- VMG 9y ago
  And yet there was a protocol change to fix the application failure.
  - tscs37 9y ago
    There wasn't a protocol change either. The hardfork that came after the burning failure of the DAO moved account values from the DAO to a new contract. The protocol remained the same.
    - splintercell 9y ago
      A hardfork is a protocol change.
      - tscs37 9y ago
        Not really, a hardfork is a explicit change in consensus, everyone decides that despite the code saying A, we do B. The underlying code is overridden for a single block before reasserting it with full authority but it does not affect the protocol. The clients and nodes in the network communicate and operate just as before, in fact, a client without the hardfork code can still receive and process the hardfork block but reject it due to not being compatible with it's consensus machine. However, this is one layer above the protocol that makes up the Ethereum network. The protocol is merely interested in the validity of a block, not how to achieve this conclusion. If you wrote a client that simply accepted all blocks as valid, it would accept this hardfork without problem, sans not having a correct balance record anymore. What it would not accept is one of the newer hardforks because those actually changed the protocol and EVM to some extend, making it not understandable to old clients. Those could be interpreted as actual protocol changes.
max_ 9y ago
Just because they hacked your webapp with XSS does not mean that the whole AWS is insecure.

verroq 9y ago

It wasn't a bug with the protocol though. It was a bug with the coding of the contract. However there has been a few forks to mitigate DDoS attacks.

gnaritas 9y ago
No, it was worse; it was a bug in the community that proved people aren't really willing to abide by smart contracts when they feel they lost. That incident split the community in half with one side willing to tear down the entire thing just to make sure the thief didn't get away with it. And now there are 2 Ethereums.
- codingmyway 9y ago
  Ironically that might be a reason for the later increase in the price of ETH compared to ETC as corporate types like the idea of being able to 'roll back' the chain in the event of fraud or a serious bug and don't appreciate that the DAO fork was a one-off event where the majority of the stakeholders were in favor due to their foolish purchase of DAO tokens.

tscs37 9y ago

DAO wasn't a protocol anymore than a Bitcoin Client is part of the protocol. It was an application that failed.

VMG 9y ago
And yet there was a protocol change to fix the application failure.
- tscs37 9y ago
  There wasn't a protocol change either. The hardfork that came after the burning failure of the DAO moved account values from the DAO to a new contract. The protocol remained the same.
  - splintercell 9y ago
    A hardfork is a protocol change.
    - tscs37 9y ago
      Not really, a hardfork is a explicit change in consensus, everyone decides that despite the code saying A, we do B. The underlying code is overridden for a single block before reasserting it with full authority but it does not affect the protocol. The clients and nodes in the network communicate and operate just as before, in fact, a client without the hardfork code can still receive and process the hardfork block but reject it due to not being compatible with it's consensus machine. However, this is one layer above the protocol that makes up the Ethereum network. The protocol is merely interested in the validity of a block, not how to achieve this conclusion. If you wrote a client that simply accepted all blocks as valid, it would accept this hardfork without problem, sans not having a correct balance record anymore. What it would not accept is one of the newer hardforks because those actually changed the protocol and EVM to some extend, making it not understandable to old clients. Those could be interpreted as actual protocol changes.

max_ 9y ago

Just because they hacked your webapp with XSS does not mean that the whole AWS is insecure.