Ah, so we're talking about two different things. Sybil attacks don't imply incompatible consensus rules. I was talking about the lack of incentive a miner has to enforce a contentious hard fork because it would erode all trust in the system by creating multiple valid ledgers.

You're referring to specific double-spend attacks. Of course a miner will double-spend if he can profit by doing so. That's basic incentive. That's very different than miners intentionally running multiple incompatible versions to provoke a chain fork.


Non-mining nodes reflect the interests of non-mining users, serving as a check on the power of miners. By not trusting mining nodes and enforcing the protocol they are integral to the integrity of the p2p system. Without them, miners have simply replaced central banks -- they may be competing against each other, but they can also collude against the userbase. And if nodes are sufficiently centralized, miners are not the only concern; it becomes much easier for other actors to coordinate Sybil attacks (tx censorship, etc).


Sure, I don't rule out a bit of that. You could have said the same thing after the XT release -- fast jump in XT nodes, which quickly fizzled out to nothing. "Thermos is such a dictator, of course people are firing up Classic nodes!" isn't a measurable phenomenon, nor is it provable. Probably better to keep other possibilities in mind, especially considering this is basically a replay of the XT release.