An inquiry into the nature of the emergent consensus of the system we call bitcoin.

I like the approach of focusing on factual technical aspects, this provides a good neutral ground that nurtures other harder subjective and predictive questions.

Regarding the questions posited towards the functioning of the BTC network, I as well am not 100% sure on the inner workings, but the general idea that 27offsuit posits is what I understand ass well. The exact workings are something that I will only truly grok once I run my own node, but for now we can investigate superficially together.

I will expose my current understanding on the matter, make note of the questions, and later consult sources.

1- Every 10 minutes, every mining-node starts computing hashes, looking for a nonce that hashes to anything below the difficulty.
2- Once someone finds such a node, they announce it to the network.
3- The transaction-receiving-nodes , check the new block, to verify if their address is credited with any incoming BTC.
4- Either the miner node, or a transaction-mining-node that has been deputized by the miner node, checks a list of pending transactions, and if the reward is worth it, checks whether they prove that the transactions were initiated by the owner (and possibly distributes a new secret to the sender-node), and advertises the results. I have some ideas on how this could be done, but I'm not sure.
5- Difficulty is adjusted, possibly by a prepublished algorithm using the diff between the timestamp of the new block and the last one.

Repeat.

Verification sometimes taking a few blocks is probably so that miners of a single block can't claim that they are being transfered ridiculous amounts of money, the next block will just fail to verify that and will continue from the previous block.

Possibility for 4.
- Owner of new block plays crypto challenge-response with nodes to verify ****. For example:
-- Talk with sending nodes to verify that they authorized the sending transactions in the pending queue.
-- Publish the authorized tx in the new block.
-- Respond to verification requests from existing nodes?

Another question that didn't quite connect is how the nodes broadcast to each other, if they do at all. They may be following the block chain, and the address and port of the nodes may be published in the blocks. So verifying nodes would never be broadcasted so to speak.

I don't think there's any deputization, the only ones who can complete transactions, and receive rewards for doing so, are miner blocks. So mining and verifying are the same thing in that sense, but other nodes can also participate in the verification by doing transactions, as they must chose the block head, the must necessarily be following the blockchain and "verifying" everything, asking only the node that published the new block to publish and "verify" their transaction.

Unless I made a mistake, that should answer most questions. Including:

What is the nature, role, and relationship between miners and nodes? miners are nodes, they get rewarded for PoW and for verifying transactions, not all nodes mine, nodes may transact or passively observe as well.

And who decides what 'valid' means in regard to confirming the work? Diff hash solvers verify what the valid transactions are and what the previous block head was. The next diff hash solver verifies the previous diff hash by choosing it to publish their new block.


That's a great start. But do you mean to imply they are necessarily different or not the same entity, etc.? They are the same entity in that they are all of the same class, everyone has the same rights, but whenever someone mines a block, they get to play the role of the authority for 10 minutes. The only difference between a miner and a non miner occurs when they get a reward, otherwise they are just another nobody. There is no verifying node, and there is no transaction-miner, there's only miners who create a new block, and transactions who request the miner to add their transaction to the block, and observer nodes who follow what's going on without any stake.

" How can this be? Quoting Satoshi: "The nature of Bitcoin is such that once version 0.1 was released, the core design was set in stone for the rest of its lifetime.": In a sense it may be right, because the code that everyone is running was decided beforehand, you cannot modify it afterwards, take the algorithm that defines the difficulty hash, it cannot depend upon what the miner publishes, as they could just end the chain by publishing nonsense, all nodes need to agree on what the new difficulty is exactly independently, otherwise values near the limit could split the network.
That said, this was probably written before forks, which through consensus achieved bipartisan code changes.

new question: How does the network respond to 2 solves within the same timeframe. Possible response: The network will probably split, possibly geographically due to latency, but eventually a majority is reached.
But what happens if there is a tie and the network is irreversibly split? Suppose that one half believes they have a 51 against 49%, and the other half believes they have a 49% against a 51%.I know this might not make sense, since the disputes are resolved on an integer scale, since each vote is essentially a 10 minute block, so a 51% vs 49% would look like a dispute that spans over 1000 minutes! I don't see what guarantees that one half of the chain eventually desist. I say chain and not nodes, because I can see a node splitting their mining across two heads if they remain neutral and hedge their bets.

Your definitions are inconsistent. We can lay down commonly use terminology, it will take a bit. bitcoin was proposed with a whitepaper to explain it. This is one of the sections:
https://bitcoin.org/bitcoin.pdf

maybe you can ask me to clarify on certain words, and I will do so, and we can forego the unproductive question on whether I incorrectly defined them or you incorrectly interpreted them.

"Nodes always work on the longest chain"

That probably solves the question on how to discern two blocks with similar timestamps. If the network is split temporarily into two, the half that solves the diff hash first will be considered and the other one would be discarded. Imagine how unlucky that would be! To solve the hash at the same time, and then lose the coinflip over whether the next hash is solved on your block or your competitor.

Seems like I was pretty spot on to me.
Here's a rough equivalency table if you are still confused.

Whitepaper's 1 - My 4
W2 - My 4
W3 - My 5, my 1
W4 - My 2
W5 - My 3
W6 -

We need to be simpler than that. And we have an agenda so we can decide to use certain definitions that others may or may not approve of.

Here the distinction I want to point out between miners and nodes is one that isn't in the whitepaper.

In the beginning it made sense that everyone using bitcoin would be both a miner and a node. But the economics and game theory are such that the natural evolution is that some nodes don't mine, and some miners don't run a node.

This is a good check to see if someone understands bitcoin beyond just googling it, because this isn't really explained in the whitepaper (is it?).

Our general or vague definition of a node is a software entity that has a interest in validating transactions, or blocks of transactions, based on the code and rules that the software entity runs.

A miner trades computational energy for a probabilistic based chance to receive bitcoin by brute forcing the answer to a computationally hard problem which is cryptographically tied to the solution provided by the last winner of the bitcoin auction .

This was written in 1999 by Nick Szabo:
https://www.fon.hum.uva.nl/rob/Cours...intrapoly.html
Here he cites two of satoshi's stated sources for creating bitcoin:



Szabo's is a polymath that has a brilliant blog of what he refers to as 'concise tutorials'. Here he formalizes a special observation:

He proposes that "..there is no algorithm running faster than q(n,k)p(n) which can invert f for more than a negligibly small number of values"

What he is starting to do I think is define a function that standardizes the computational power of a hardware's ability to brute force.

The conclusion tho is very important imo:

What his conclusion expresses is the concept of an "asic" bitcoin miner:

Szabo's formalization is important because it shows that there IS in fact some formalizable computational limit to the total computing power of a network. Theoretically in principle it could be throttled by a math rule. Most people credit Satoshi with this discovery. It was a less of a discovery and more of a formalization even as far back as 1999.

People often say that Satoshi solved the byzantine generals or consensus problems. Probabilistic Byzantine Consensus is what Nick Szabo defines 'nakamoto consensus' as:

It's reasonable to consider Szabo is referencing this paper from 1997

https://people.csail.mit.edu/silvio/...0Agreement.pdf

The problem.

"...up to 1/3 of the generals including the commander in chief-may be traitors.... traitors may represent their orders differently to different generals, they may not send any information to someone,and they may claim to have received nothing from someone else....they need is a way to exchange messages so as to always reach a common decision while respecting the chief's order, should he happen to be honest. They need Byzantine agreement.



Byzantine agreement.

"...we must first sketch its classic underlying communication model...Modernizing the motivating scenario a bit, generals are processors of a computer network....Every two processors in the network are joined by a separate communication line, but no way exists to broadcast messages. (Thus, though a processor can directly send a given message to all other processors, each recipient has no way to know whether everyone else has received the same message.) "

That paper formalizes the problem of nakamoto consensus. But it also provides the solution:

The basic idea is when looking for consensus on valid agreements, the introduction of a probabilistic competent into the validity rules, sets a nice limitation to any 'dishonest' or 'counterfeiting' nodes trying to corrupt the consensus.

Note the authors foresee their solution as the birth of an important technology:

Satoshi isn't the one that solved this problem. He implemented the solution.

Was this thread started with a question in mind? I am not faulting anyone in the thread, but I found the whole thing -- unappealing to read. Is the question what is the difference between a miner and a node in the BTC network? If so, this seems more like an engineering question than a economic one.

Nash defined Ideal Money as being pegged to something cost invariant to produce.

If you were Nash/Satoshi how, ideally, how would you constitute adjusting the cost to mine each bitcoin invariantly when you can't know or predict the total computing power that will be mining it each period?

Is that a separate question from economics?

Well, apparently Nash isn’t as smart as you think that he is.

In the current system, dollars are cost invariant to produce…

Why does everyone want to point out he is dumb? Its astounding
The goal was to find an apolitical basis OTHER than of the whims of a government. So you need a throttle. Cost invariance is a nice throttle. If its not cost invariant then the significance bestowed on it will have geo-political ramifications and supply shocks etc....



If we are to come up with a rule for ideality, giving control to a single government is the opposite of the intention...given we want something apolitical, we are looking for a commodity that can handle the burden of being 'special' (that can handle a 'gold rush').

that is not what I said


It’s not a bike. You are overthinking it.



Who said anything about government?

cost invariant money is a means not the ends. Using dollars or another nation or centrally banked currency is not ideal because it serves as an unstable basis based on the political whims of the issuing government. If the issuer were trustworthy...you would be otherwise correct...



The cost invariance is in regard to the idea of a commodity money, which could also be inherited if a money could be pegged to the price of a cost invariant resource or basket of resources in a trustworthy manner etc.

If satoshi went to nash to say what would be ideal here? And Nash said, make each bitcoin cost invariant to produce....

How does Satoshi account for all of the rush of computer architecture being created to mine bitcoin at an unpredictable and ever cheaper cost?

Knowing its got to be constituted BEFORE hand, if bitcoin is to have its covenants secured.

If anyone that feels they know bitcoin well enough thinks they know the proper technical answer....I'm going to ask of it immediately..."does that make bitcoin cost invariant to produce?"

https://en.bitcoin.it/wiki/Consensus

The probabilistic byzantine paper showed us that introducing a probabilistic component into our consensus rules gives us byzantine fault tolerance (ie up to 1/3 corruption)
https://forumserver.twoplustwo.com/s...9&postcount=11
Miners probabilistically/brute for the next valid block, nodes implicitly validate or only see valid blocks/transactions.

But what is the difference between a miner and a node? It was Szabo that predicted, thru formalization, the mining race:

https://forumserver.twoplustwo.com/s...44&postcount=9
Nowadays people run specialize mining hardware, which search for a valid block to receive the next scheduled bitcoin award, and the miners have learned to pool together, and thus they can share the same validator. At the same time not all nodes or validators mine bitcoin.

https://bitcointalk.org/index.php?to...sg1611#msg1611

Early on Satoshi explained of bitcoin:
But it was almost immediately that people saw in its current form that bitcoin didn't scale to serve the world with a globally used money. In trying to address the issue, and force the early group to confront it, Jeff Garzick proposed a patch that increased the blocksize. He suggested the community move to it before the entropy grew....



Theymos pointed out that the patch forks you from the propriety of the network...Satoshi immediately affirmed:


But Garziks stunt was an expression of the game theory he was playing out in his head...his point was made...

Yet Satoshi made the innocent statement that they can easily deal with the problem later:



And remember the bitcoin wiki says this:
https://en.bitcoin.it/wiki/Block_size_limit_controversy
So the statement of creating a flag day is said here to be a ruse.

Who ya talking to, buddy?

Its a question people want to ask and want to understand.