I am not sure, but i think you do not understand how the bitcoin network works. Every day we have several "hard forks" in the form of orphaned blocks.
Thanks for the lesson, but I do understand the consensus mechanism, and I know about orphan blocks. Perhaps I am using the wrong name, but those that you describe are not "hard forks", not even "soft forks", it is just the normal operation of the protocol. Both the orphans and the winning branch use exactly the same version of the software, and the UTXOs in the orphaned blocks become invalid by definition.
What I am discussing is a fork of the blockchain where each branch uses a different protocol, and every message is identifiable as belonging to one branch only, and is accepted (or seen) only by clients and miners who are running that version of the protocol. Blocks before the split are valid under both protocols, but blocks after the split are valid only according to their specific version. Each transaction request is directed to one branch only; if a client wants to move both clones of his coins to the same address in each chain, he must issue two requests, and they may or may not be accepted independently.
you insist that A and B blockchains can coexists - with the same mining power behind them!- for more than say 2 consecutive blocks (on avg 20 minutes), and that they both individually represents bitcoin as per se. No, they do not.
No, that is not at all what I am discussing, see above.
It may be possible to have some "merged mining" scheme where a miner can attempt to mine both chains at the same time. But I am not considering that possibility. I am assuming that each miner chooses only one of the two branches to work on, by upgrading or not to the appropriate software and/or talking only to the appropriate set of relay nodes.
After a hard fork, there is no automatic "synchronizing" of the two branches (and re-joining them would soon be impossible, even with substantial hackery).
Your B chain will not ever propagate because: the moment you fork, you also inherit the difficulty of the A chain. That difficulty is so high, that your laptop will practically never finds the next block
Yes, as I said, a hard fork that cannot count on attracting most of the hashpower right away (like that kid's) would also have to lower the difficulty temporarily to ensure a fair block rate, until the automatic adjustment can take over.
But a hard fork that can muster 25% (say) of the hashpower perhaps could afford having 1 block every 40 minutes for a couple of weeks. So the difficulty adjustment may not even be necessary.
if you think that the state of the blockchain is a political decision, you should STOP posting bullshit here please. You do not understand the very basic idea: consensus through mathematical proof independent of central authority ( which is politics...)
At the OP_MUL fork, everyone choose to switch to the new protocol, recommended by Gavin, even though it meant "rewinding" the blockchain and invalidating some deeply-confirmed transactions. If tried to do a hard fork of my own, surely no one else woud switch. The difference between the two cases is not in the protocol or mathematics, it is politics.
The good working of the protocol depends heavily on the behavior of its human players, about which nothing can be proved mathematically. It assumes that a majority of the players will be driven by immediate greed, and therefore would choose alternative X over Y when considering many specific choices that have been considered when designing the protocol. However, that assumption is not certain, and the possible situations and choices that the players may face is not bounded.
We are already seeing some situations and choices that may not have been foreseen. For example, the transaction fees being insignificant compared to the block reward, and some powerful miners apparently choosing to mine empty of half-full blocks rather than full ones.
Poll
