Bitcoin is unlike other OSS in that one simple bug can be catastrophic. We should try to decrease the attack surface (# of Lines of Code), not increase it.
-
-
Show this threadThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I disagree on this, a spec implemented in multiple clients would be more resistant. But it's hard to do for Bitcoin as the reference implementation has a lot of implicit quirks that the networks relies on: those would need to be made explicit in the spec.
-
A failure in consensus code would not make the network more "resistant". Rather, it might accidentally split the chain or partition the network. That would be disastrous.
@adam3us -
But a bug in one of the implementation would have less impact than a bug in the sole reference implementation.
-
Debatable and data not on your side tbh. We will have more empirical evidence as time goes on. Bitcoin consensus failures so far: 0. Ethereum: multiple & will only get worse as the client gets more complex.
New conversation -
-
-
also multiplied complexity of consensus bad interaction. ethereum had multi-way consensus failures across it's implementations. consistency is more important than pedantic correctness where it does not violate expectation.
-
Agreed. Anyone who has written software to specs know that how hard, if not impossible, to precisely match specs to the letter (assuming specs are perfect to begin with, which they are not). In those situations, maintaining consistency is critical to the health of the network.
-
even that consistency might be "wrong" from specs' PoV.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
