| >You either have to ask people (i.e. you need a priori knowledge obtained out-of-band) Again, the only reason blockchains need consensus is to allow people to interact with each other - consensus is between people. Computers are just tools to make that easier. It's a fundamental contradiction to assume you can use any blockchain to make any economic transactions without interacting with other people - because economic transactions require other economic entities. Of course when you assume something false you can prove any absurd result, like that PoS wastes same resources as PoW. PoW relies on social coordination in the short term, because short term attacks are cheaper, so in the case of a 51% attack people would have to organize fast. PoS is extremely safe in the short term, and only maybe falls back on social coordination in the long term (again, only in the case of an attack), which is the correct security model. >deterministically choose the fork that the economic majority of people use (which is the problem PoW solves) No it doesn't. Mining revenue is an insignificant part of what the real consensus in any PoW coin is. For a while BCH had biggest revenues after the fork (because of their difficulty algorithm). Ethereum has higher mining revenues than bitcoin for months now (last 24h: $49M ethereum, $31.3M bitcoin) - does that make ethereum the true bitcoin now? |
Did I say otherwise?
> Of course when you assume something false you can prove any absurd result, like that PoS wastes same resources as PoW.
Well, no widely-used PoS system exists (so we have no real-world examples to learn from), but despite this, you're insisting that no PoS system will use more than PoW from now until the last blockchain goes offline, despite these systems (in expectation) driving essentially unbound amounts of revenue. That's quite an extraordinary claim!
Let's steel-man this. Let's assume that a PoS blockchain becomes so widely successful that its token becomes a major world currency. Then what? Controlling a PoS node would be like controlling a country's reserve banks and mints. So, what keeps these nodes safe from asshats breaking into them and using them print themselves money? Like, why can't an armed band of asshats show up at my server rack and physically steal my validators' keys?
The answer of course is that the building security and law enforcement officers keep this from happening. But, where do these people come from? Who pays them? Where do they get their equipment? What do they do with the asshats they catch? How do they deal with escalations from asshats, and stay ahead of the asshats' tactics? How much energy is going into keeping these PoS nodes secure?
It appears that there is energy involved in keeping the PoS system running in the face of asshattery, and that energy is proportional to how important it is that it remains usable for the societies that rely on it. It seems, then, that the more successful PoS becomes, the more it co-opts the very infrastructure that keeps today's financial systems secure. That's a lot of energy!
So, in the event of success, I have no reason to believe that PoS will take less energy to secure than PoW, once I think about what has to go into securing a successful PoS system. At least with PoW, I can rest assured that if the asshats hijack a mining rig to print money, they'll have to continuously out-mine the rest of the world in perpetuity in order for their coins to remain realized on the canonical chain. PoS doesn't have that resiliency, which necessitates building and maintaining an extrinsic security apparatus to keep the staked coins from getting stolen in the first place. This security apparatus -- including all the laws, supply chains, manufacturing, and so on to keep it going as it becomes a more and more valuable target to asshats -- is on the MC side of the equation.
> No it doesn't. Mining revenue is an insignificant part of what the real consensus in any PoW coin is. For a while BCH had biggest revenues after the fork (because of their difficulty algorithm).
You've completely misread my comment. Miners mine on the chain that is most profitable to them, and the blockchains they mine on encode the history of their activities. Even though during a chain split it's not immediately apparent which resulting chain will attract the most miners over time, it does become apparent quickly enough. The revenues (and thus profits) come from users actually demanding the coins.
> Ethereum has higher mining revenues than bitcoin for months now (last 24h: $49M ethereum, $31.3M bitcoin) - does that make ethereum the true bitcoin now?
I thought it was widely understood that Bitcoin and Ethereum are not the same thing? If there is contention between two forks of the same blockchain, then PoW provides you a way to determine which one has more demand. PoW doesn't tell you anything about two different blockchains with two different difficulty algorithms (but it might tell you something about two different blockchains with the same difficult algorithm, such as Bitcoin vs Bitcoin Cash).