Quote from: 2112 on May 14, 2017, 07:33:44 PM

Quote from: dinofelis on May 14, 2017, 04:10:39 PM

I think my argument is quite watertight concerning the non-power of full nodes, and I've never seen a LOGICALLY ARGUMENTED rebuttal to it, that doesn't confuse full nodes with users, or doesn't confuse it with the power game of a hard fork.

You've spent too much time arguing with dumbasses and marketroids which caused you to start believing in your own bullshit.

Almost from the inception people understood the importance of distributed node network, even if under centralized control:

1) ability for nearly anyone and relatively cheaply audit (in the accounting sense) the totality of the system.

2) distributed node architecture makes it easy and cheap to hide the actual expensive nodes (either due to mining or transacting/exchanging) from both attacks launched over the network.

3) not currently implemented by Core, but possible, the broadcast architecture makes it possible to use non-Internet methods of communication with super-extreme bandwidth asymmetry (high download but little upload).

1) audit, yes, I do not deny this. You can find out for yourself whether the bitcoin block chain is doing what you thought it was supposed to be doing (that is to say, whether it corresponds to the protocol programmed in the node software). Now you can learn whether that is a yes or a no. But that's about it.
It is like checking a signature. You can see whether it is right or not. But that's about it. You can't DO anything about this. You can write an article about it here on the forum, or you can write this on coindesk or something. "hey, guys, the unique block chain is not checking on my node !". Duh.
You may use that knowledge to sell your coins (using *another* node that will allow you to send transactions, and does accept the block chain's protocol!) because you may think that bitcoin is screwed now. But that's about it. Yes, it will INFORM you. No, it won't do anything about it.

2) how's that ? We know their IP addresses. Why would an attacker bother to attack Joe's node in his basement, when he knows the node IP of BitFury, say ? But I will agree with you that a P2P architecture has some network advantages if direct internet connections are problematic. However, you don't need the BITCOIN P2P network for that. You can use just ANY P2P network, like Tor, to do the network communication between the customer (the wallet user) and the provider (the miner pool backbone). You can argue: yes, but what if THEIR backbone fails ? Well, in as much as the bitcoin P2P network finds a way around, just ANY internet connection will find a way around.

3) actually, that would rather be the direct connection between user wallets (customers) and the miner pools (providers). Note that these miner pools may just as well have a DISTRIBUTED network of proxies (distributed, but centralized under their control) in the same way that google and facebook have distributed but centralized proxy servers all over the world. What full nodes do, is to provide miners with a free proxy service.

The point is that non mining full nodes don't contribute to the DECENTRALIZATION (not to the distributedness) of the network. Decentralisation is about power and politics ; distributed is about network architecture. full nodes don't have the slightest bit of political POWER.

Quote

Dinofelis, you clearly seem to have some academia background. But it must have been some theology school (or something equally anti-scientific) to be so unwilling to actually refer to the available references and verify your own assumptions.

This is funny, because I'm the only one never to use something else but logical arguments, and try to reason from a given to a result, without thinking about politics, autority, "others say so", etc... but just sheer reasoning. I'm totally a scientist. I'm actually the one thinking I came in some sort of church or something, where established dogma may not be questioned. One of these is "full nodes are important to the decentralization of the network".

Now, given the fact that this in my honest opinion totally flawed statement is so often used to take decisions, I would think it is important to establish its veracity or its falsehood. You would think that if Pythagoras' theorem is used all over the place, that *there's a solid logical proof* of it. But no. If presented with a proof of the opposite, there's no watertight demonstration of the veracity of the theorem in sight. Only "you're wrong" or "you're a shill" or "you're religious" or whatever. No mathematician ever would say, to someone that says "I think Pythagoras' theorem is wrong", "you're religious". He would just give the demonstration of the theorem. That should convince. I think I have enough arguments to show that full nodes don't contribute anything to political decentralized power. Show me the proof of the opposite.

For that, I discussed elsewhere the following Gedanken experiments, which will abstractly show one or the other: full nodes impose their view on miners, or the other way around.

Starting condition: today. Miners and nodes agree on the same, current, protocol. I consider two (abstract) cases to come to the conflicting situation, where miners and full nodes are in disagreement: A: the miners change something ; B: the full nodes change something.

A) suppose that the 20 miner pools (>99% of hash power) decide amongst themselves, to start mining blocks of 1.2 MB. They publicly announce this, and there's also a version of bitcoin node software that can do this. No non mining node agrees. They all stop. The miners continue making the block chain with 1.2 MB blocks. What happens ? Do users connect their wallets to the miner nodes, and continue using bitcoin, or does bitcoin come to a grinding halt, no exchanges, no transactions, nothing, all the time these miners keep making the new chain (and no old protocol chain is available) ?

B) suppose that the 20 miner pools keep on working on the bitcoin chain as today, but ALL non mining full nodes want to impose blocks of 0.5 MB as from tomorrow. For one or other reason, all full nodes agreed upon that. So they all implement the new software, that doesn't agree any more with the block chain the miners are making. Do users connect their wallets to the miner nodes, and continue using bitcoin, or does bitcoin come to a grinding halt, no exchanges, no transactions, nothing, all the time these miners keep making the old chain (and no new protocol chain is available) ?

From these two exercises, what can we conclude about the political power of full nodes over miners concerning bitcoin's protocol ?