Will mining centralize around one single miner? No one knows. This is why Bitcoin is still a risk.
But if this doesn't happen--if there remains more than a single miner--then the fee market does exist.
I think the point of contention is really about this conclusion.
My understanding is that it may also happen with several miners if they have an incentive to synchronize their mempools (with a mechanism like IBLT).
I ask, just because I want it to be clear to readers that this "point of contention" is about a hypothetical future scenario. And that we already know that there are hypothetical future scenarios, such as mining centralized around a single super pool, that would be bad for Bitcoin.
It absolutely doesnt. Miners already optimize for profit by centralizing in various ways. The recent SPV mining near-catastrophe is an excellent example.
Your conclusions hold under no existing and future scenario. It makes assumptions that are untenable and require absolute altruism from miners.
The crux of the issue is that miners incentives are not aligned with the network's users. The max block size cap mitigates the fact that they are expected (for security reasons) to prioritize financial profit over network decentralization. Yes, there might be no other way to align these incentives than by forcing it through consensus code.
hey! i disagree!!
in my view miners are nothing if it was not for the users they validate the transactions for: users do the speculation, users gives bitcoin its value, which then makes it profitable to mine (and not any other altcoin) .
remember what happened btc price wise when ghash.io almost had 50% of the network? massive sell out..
hence, miners do have an incentive to keep on mining the longest VALID chain, which, by definition means the one that has the more value potential.
so they better not screwing around with some power grabbing fork, else they loose everything...
this is bitcoin's consensus.
I agree with most of what you said but I don't believe it really addresses my point.
The GHash situation is very different as it was very public. The SPV mining fiasco is a better example of a situation users were not readily aware of yet causes considerable risk to the ecosystem. The miners do have incentives to maintain the users trust but they have a certain flexibility and ability to optimize for profit in a "bend but don't break" manner. Anyway, Peter Todd always makes a good point that we should not build this system based on "best-behaviour" assumptions but rather expect any potential attack.
The centralization I refer to relates to geographical and technical optimization that give larger miners a slight advantage over smaller ones that slowly but surely dries out the smaller miners revenue.
It is not something that is immediately apparent to most users and the extent to which such centralization can occur is difficult to examine.
That is not to suggest they are expect to fork off or create immediate damage but it is a slippery & dark road that can slowly lead to a level of centralization that enables considerable throughput hence the centralization pressure externalized to the network nodes.
