Sitemap
Thεο ¯\_[ôo]_/¯

Thεο ¯\_[ôo]_/¯

Interested in Freedom, Decentralization, Crypto.

Talking Nodes with Marlow

6 min readOct 9, 2017

--

I discussed Bitcoin Nodes with Marlow (aka, @profFaustus, CSW, Satoshi, whatever). The conversation is disjointed, as it was made in a series of comments that are hard to follow. I believe this conversation has value, so I am including the conversation in its entirety here. The conversation started from my reply to Adam Selene’s post on Nodes — which goes as:

ThεοδοsιοΣ
Jun 15

Hogwash — absolute hogwash.

A mining node cannot distinguish if it is receiving validation information from another miner, or from a fully validating node. A mining node does not do any extra checks to verify if a block broadcast was sent by a mining node or the nearest validating node. So if it receives a valid block signal from a validating node, it treats it the same as if that information came from a miner. The only node can be called a miner is the actual node that solved the POW. All other mining nodes are simply ‘nodes that did not get the block’ and they appear no different than any other full validating node.

Sure, a full validating node does not ‘create history’, but if it detects invalid blocks/forks/txs, it will disconnect from the ‘offending node’ as will all other nodes, in effect isolating the node leaving it with no ability to broadcast its block to the network. When a full validating node disconnects from another node, it also doesn’t know if it is disconnecting from a mining node or a non mining node.

Disconnecting from a node that broadcast an invalid tx/block/fork is not doing nothing.

Beyond that, the actual ASIC chip on a mining farm that solves the block is dumb to the network and broadcasts its find to the farm’s node, which is a full validating node, which then broadcasts it’s find to the nearest nodes.

Absolute hogwash

Marlow

What a shame you don’t understand how the network works. You are describing a power law Meshed network.

Bitcoin is formed as a NSW Random graph with a distance of approximately d=1.32.

In the paper, On red balloons and bitcoin [1] it is noted that Sybill resistance cannot be achieved At a distance greater than d>4.

So your idea of transactions and blocks hopping across the mesh is inapplicable. It does not matter if 10,000 wallets decide they want to block transactions or blocks within bitcoin. The mining network forms what is termed a giant node. There is no such thing as a full validating node. At best, each of these nodes stops broadcasting. As it prunes itself from the network, the rest of the network actually acts to broadcast more efficiently.

All that matters in bitcoin is mining. The more full validating Sybills you add the more you get attackers to waste their money.

Your understanding of the mining process and the network connectivity equate to about zero.

[1] https://arxiv.org/abs/1111.2626

ThεοδοsιοΣ

Interesting ‘bait and switch’.

In case anyone actually reads the paper (no paywall here: https://users.encs.concordia.ca/~clark/biblio/bitcoin/Babioff%202011.pdf), the paper made suggestions in 2011 based on network nodes all mining. The bitcoin network has evolved many iterations since then.

More to the point: You know you are wrong, and I know you are wrong, and you know I am right about my previous comment. Your game therefore is to convince people that do not understand the network topology of this false narrative. A mining node that broadcasts to the network with a block that contradicts consensus will isolate itself by doing so and then not be able to propagate blocks — this is what keeps miners ‘honest’ as honest is not a term of endearment here, it is a technical term.

I digress. Consensus is not just mining. Miners participate in consensus, they don’t drive it.

Your ‘bait and switch’ will no doubt fool some.

Marlow

And I suppose Wallets signal some how?

ThεοδοsιοΣ

No. And I am talking about Full Validating Nodes, not wallets on mobile phones, browsers or even wallet only clients. ASICs use FVNs to propagate blocks to the network. An FVN, even one connected to a miner, does not know to distinguish if it is receiving a newly propegated block from a FVN connected to a miner, or if the FVN it is connected to is receiving the propagation some hops away from the miner — it doesn’t even perform a check of any sort. At any one instance of a block being found, it is one miner that found it, and all other mining nodes are no different in appearance on the network as a FVN. This is not rocket science, it is very basic to understand.

Marlow

Wallets = Full Validating Nodes…. Learn to stop the 1984 double talk.

Get Thεο ¯\_[ôo]_/¯’s stories in your inbox

Join Medium for free to get updates from this writer.

The nodes used by a miner has multiple parts. You should learn that a node is a super-system and not the component.

And yet you still fail to understand… Just to move goal posts and equivocate.

ThεοδοsιοΣ

Dumb ASICs (dumb to the network) connect to the network via Full Validating nodes. ASICs solve the hash, they route tot he network via a FVN. The Miner (with the ASIC that solved the hash) then broadcast their finding to the network. The miner actually presents this block as a valid block that is the tip of the longest chain. The miner’s node then checks that there is no conflicting block form other nodes, (this does not mean that a FVN pushed different information, I never suggested a node pushes anything).

Apart from the dumb ASICs and the miner’s FVN, the other parts you describe in your super-system are just routing equipment and a controller.

Now if you like to use different terminology, that is semantics.

But I digress again, i know you know this, this is simple and doesn’t need explaining to anyone technical. But you are deceiving people that do not have technical knowledge.

You have not addressed the point I made repeatedly:

We all know that miners are listening to the network to make sure a valid block has not been broadcast before them. If they receive this info, they simply start mining the next block in a race, right! simple 101 — so, again!:

How does a miner know to distinguish if this information is coming form another miner, or from another full validating node that is a few hops away from the miner that broadcast?

If you answer this correctly, your whole argument falls to pieces and you lose face. Go on, explain if the miner can distinguish a mining node from a non mining FVN.

Marlow

ASICs plus other components are Nodes.

You are discussing wallets. Systems used by users. These do not push to miners. You should do some checking on what the network graph in Bitcoin is structured as…

Then, I do not think this in in your capability…

ThεοδοsιοΣ

  • “ ASICs plus other components are Nodes.”

That is not a technical fact, that is a deceiving narrative.

Nodes can operate without ASICs. You can run them with an ASIC, or with a CPU or GPU/s (although not competitively any more, but certainly when the WP was penned) and you can run a full validating node without a mining rig. It still seeds the network, validates, relays etc. User wallets do not.

But you know this, so you are just pushing a narrative, a false one.

When a mining node receives relays from the network it does not distinguish if the information is coming from another mining nodes or a FVN, because it can’t. So whatever else you are peddling on top of the narrative you are pushing is moot.

(Then from left …)

Jonald Fyookball

i’ve been following this conversation and I read the paper referenced. I do not think your point that Fully validating nodes somehow helping honest miners (or punishing dishonest ones) makes sense.

As the paper says, bitcoin assumes nodes can take a false identity. Since nodes can take a false identity, they can continue to try to DDOS other nodes. Others disconnecting from them won’t stop that as they can switch identities. It’s not like nodes ask other nodes what nodes are connected to them and then make decisions, I mean come on… if a node keeps sending my node nonsense, I can block that node. End of story.

ThεοδοsιοΣ

I never suggested nodes can punish a miner. If a mine broadcasts an invalid block, the block won’t make it through validating nodes with the effect of a miner isolating himself off the network, as if the block never existed. That is how p2p works and you have said as much in your closing sentence, all validating nodes will reject gibberish and not pass on the broadcast. The miner’s block becomes like the tree in the forest that fell but there was nobody there to hear it.

--

--

Thεο ¯\_[ôo]_/¯
Thεο ¯\_[ôo]_/¯

Written by Thεο ¯\_[ôo]_/¯

Interested in Freedom, Decentralization, Crypto.