Quote from: JorgeStolfi on March 02, 2015, 11:09:29 PM

Quote from: oda.krell on March 02, 2015, 06:53:34 PM

This security is provided in the form of total hashpower that can be assumed to be honest (an assumption that is based on economical incentive, so it's a rather safe one).

Well, I find that argument quite unconvincing. As you pointed out in the case of day trading, one cannot assume that every player will just try to maximize the amount of coins that he gets from the next block. Large corporate players may have more complex, longer-term goals, and motivations other than greedy coin-slurping.

Counter question: "Why do most people (mostly) trust their governments?"

In a country with (by and large) working institutions, the government's incentive to act responsibly, all other things (like "a feeling of responsibility") ignored, will be to reelected. That is certainly not an incentive that we can be sure to dominate all other possible incentives, but it's a pretty strong one, as evidenced by the efforts politicians go through to be reelected. Unless presented solid evidence to the contrary, we can assume that incentive is good enough. From my own experience: in the countries I have lived in myself, I have at most seen evidence for imperfect organization, "stupidity", not a fundamental breakdown of that "democratic incentive". I'm sure there are countries where that's different though.

Anyway, a similar argument applies to the Blockchain: we know there is a strong economic incentive for miner's to be honest (the value of their hardware investment would be drastically reduced if the Blockchain is considered compromised, in addition to the substantial loss of value of any BTC position they might be holding). Any stronger incentive that could override the previous one is hypothetical so far. To conclude that the second incentive matches the first one requires additional evidence, because we have extremely strong evidence that the first incentive "works" (Blockchain uncompromised, no known double spends, for example).

It's the choice between believing something that we know to exist (and matter a lot) to govern a system, or believing something that hypothetically could exist (and matter) to govern it.

Rational people should pick the first option as the most likely factor to govern the system, while keeping their eyes open for evidence for the second factor. I know I do.

Quote from: JorgeStolfi on March 02, 2015, 11:09:29 PM

Quote from: oda.krell on March 02, 2015, 06:53:34 PM

Your 'Series B' blockchain is either indistinguishable from the original one, in which case it doesn't divert any value, or, if it is in any way distinguishable, it will be ignored by the majority of miners, so it cannot fulfill the above security critical functions.

It may differ from the main chain, e.g. in block frequency, block reward schedule, mandatory minimum fees, etc.

In another thread, I discussed a scenario where miners conspire to force a change in the reward schedule, postponing the next halving for 2 years (without changing the limit on total coins). Some argued that long-term holders would vehemently oppose that change because it would mean higher inflation in those two years, and they would rather do a hard fork of their own. But then, going back to that kid's fork: if he changed the schedule to bring the next halving forward a yar, then those holders would surely love his Series B coins, no? Since those holders would automatically have as many Series B coins as they have in Series A, why not value and use them too? Wink

To make the "time travel" advantage more plausible, suppose that a hacker manages to steal the 94'000 BTC from the USMS wallets. Six hours later, after meeting in some undisclosed location at the invitation of some undisclosed entity, the CEOs of Coinbase, Kraken, SMBIT, COIN and a few more US bitcoin companies, as well as the president of the Bitcoin Foundation, issue a joint announcement: "for the future of Bitcoin", they are starting a hard fork that branches off just before the heist, and will henceforth work with that fork only; so all clients had better upgrade asap. Moreover, with the help of that undisclosed entity, they are setting up a dedicated mining installation and temporarily reduce the difficulty so as to ensure an adequate block rate, but urge all miners to switch too. They will try to re-issue any canceled transactions, or indemnify legitimate losses. Naturally many bitcoiners, exchanges, and miners all over the world will revolt and will keep mining and using only the old chain, while others will use both.

Can you say that this scenario is impossible, and that one of the forks will immediately lose all its value?

You completely ignored my argument about security being the deciding factor - probably because you concluded your first point about miners' complex economic considerations took care of that one Cheesy

The "alternative blockchain" you describe (different block frequency, for example) will be extensionally different from the original Blockchain. Miners will not point their hashpower at it, thus making the "alternative" blockchain mostly useless. Any value it would hold would be comparable to that of an (completely uninnovative) altcoin, i.e. not enough to worry about.