Quote from: bargainbin on February 13, 2016, 03:19:03 PM

Quote from: madjules007 on February 12, 2016, 10:54:46 PM

Quote from: bargainbin on February 12, 2016, 04:57:48 PM

Quote from: Lauda on February 12, 2016, 04:24:39 PM

Quote from: bargainbin on February 12, 2016, 04:10:29 PM

Admitting that node count is irrelevant kills the biggest argument against increasing the blocksize Undecided

The number of nodes over a longer period of time is relevant. Node count as a metric used for 'debating' is irrelevant. You can't properly measure the actual number of nodes and the number can be faked. Your attempts at trolling won't work.

Nonsense. Number of non-mining nodes (call them what they are: wallets) is irrelevant to Bitcoin security.

Completely wrong. This implies that the entire network could be comprised of SPV nodes that trust a small, centralized group (miners) regarding the validity of the blockchain. That is not a trustless system at all.

More nonsense, it doesn't. Am assuming that your inference process works something like this:
  1. You claim that the hundreds of padlocks thrown about on our lawn help secure our house.
  2. I disagree, and point out that the padlocks littering our lawn add nothing to our security.
  3. From this, you infer that I'm saying that our house is secure without those padlocks.
That's flawed reasoning. I've never suggested that Bitcoin is either "a trustless system" or secure, merely that non-mining nodes don't *add* anything to security.

Anyone reading can see this is a ridiculous analogy. Instead of inventing an irrelevant analogy, try proving the actual statement wrong. If "non-mining nodes [are] irrelevant to Bitcoin security," that implies either that "Bitcoin is secure with only mining nodes" or "Bitcoin is insecure no matter what."

My statement above specifically addresses the former. It is absolutely not secure with only mining nodes. The burden is on you to prove the second statement.

Quote from: bargainbin on February 13, 2016, 03:19:03 PM

Quote

It means that miners can level endless attacks on the entire userbase, from double-spending to transaction censorship.

Yes, they could, if that happens to be in their calculated best interest. Non-mining nodes (wallets) prevent this only as far as being ...you guessed it, wallets. You can rely on your wallet for making sure the correct ruleset is being followed. If you chose to rely on other_nodes (SPV wallet), or other people (web wallet), that's called trust.

Yes, relying on other nodes = trust. And without non-mining nodes, there are only mining nodes. Hence "miners can level endless attacks on the entire userbase, from double-spending to transaction censorship." You say "other nodes"...but if there are only mining nodes, the userbase, by definition, trusts miners. Hence why the existence of non-mining nodes secures bitcoin: by decentralizing hash power from validating nodes.

You said "non-mining nodes don't *add* anything to security." For individual users: Anyone who transacts with bitcoin without validating their own transactions is insecure and dependent on trust. For the bitcoin system: Without non-mining nodes, miners can level endless attacks on the entire userbase, from double-spending to transaction censorship.

Go ahead and actually prove those statements wrong instead of making a bizarre, irrelevant analogy.

Quote from: bargainbin on February 13, 2016, 03:19:03 PM

Quote

This situation is analogous to banking customers (the bitcoin userbase) depending solely on central banks (miners) to uphold the integrity of the system. It doesn't work in the real world and it won't work in bitcoin.

Your level of participation/("say") in Bitcoin is proportional to your hashrate. The only voice BTC holders have is the market: You may [try to] sell when things don't go your way. Again, running non-mining nodes offers you nothing above verifying transactions for yourself (again, used as your wallet).

No, the bolded is incorrect. Hash rate is irrelevant without rule enforcement. If miners are the only party enforcing the rules, all users trust miners. My voice is my node. I enforce the rules not only to validate my own transactions but to enforce the rules of the system, helping to prevent miners or other attackers from breaking them.

The more nodes that exist that are enforcing the same rules, the more likely the longest valid blockchain data that you receive from them is accurate. If there are only five validating nodes in the network, it only takes six Sybils to make many attacks on the system easily possible.

Regarding this thinking that miners are the end-all, be-all of bitcoin:

Quote from: madjules007 on February 08, 2016, 05:43:21 PM

Miners are at the whims of nodes. Running an ASIC farm doesn't give you any rights over what node software I run. And an army of nodes enforcing bitcoin's rules (including a 1MB block size limit) is here to tell you that they do not give a shit about the opinions of miners. If there is a question of multiple blockchains, miners are squeezed by economic incentive far quicker than are nodes. You won't be squeezing my nodes.

Miners have to contend with the idea that if they cannot reach 100% consensus on a single rule set (and therefore a single blockchain in a chain fork situation) that lack of node consensus will force them to chain fork and therefore choose between two distinct blockchains. A rational miner would probably rather avoid that situation. I continue to enforce the 1MB limit -- as do others -- to force rational miners to tread carefully when considering whether to hard fork without full consensus.

Quote from: bargainbin on February 13, 2016, 03:19:03 PM

Quote

Non-mining users and miners have competing incentives. Miners are only concerned with profit -- historical attacks (withholding, double-spend, tx censoring) support that.

Yes, something that's often glossed over. But non-mining nodes are impotent -- if only [for the sake of simplifying the argument] because "bad," competing nodes could be trivially/inexpensively created.

It may be true that spinning up some nodes is trivial -- but mounting a successful Sybil attack may not be so trivial or inexpensive at all. Spinning up nodes for nodecounter.com is very different from being able to successfully attack the system. Non-mining nodes obviously aren't impotent because we are talking about theoretical Sybil attacks, not historical ones. Cheesy

Quote from: bargainbin on February 13, 2016, 03:19:03 PM

Quote from: madjules007 on February 12, 2016, 10:54:46 PM

The only checks on miners' incentives to attack other miners or users are 1) other miners (who might control enough computing power to prevent computing power based attacks) and 2) non-mining nodes (who might control enough nodes to prevent Sybil attacks).

But how do you keep missing something which seems so self-evident: non-mining nodes could be created by anyone, including the miners they're meant to keep in check.

Anyone can spin up some more nodes, including miners. So what? A miner bringing another Classic node online doesn't mean my Core node is going offline.

Quote from: bargainbin on February 13, 2016, 03:19:03 PM

Quote

By definition, nodes reduce centralization.

By definition of ...what?

Every node must receive and process every transaction in every block. That distributes blockchain data to all nodes who enforce the rules, rather than a system where nodes trust a central source.

Quote

The more nodes that exist that are enforcing the same rules, the more likely the longest valid blockchain data that you receive from them is accurate. If there are only five validating nodes in the network, it only takes six Sybils to make many attacks on the system easily possible.

Quote from: bargainbin on February 13, 2016, 03:19:03 PM

Quote

Since every node must receive and process every transaction in every block, they are the "peers" in the p2p network. They are what gives meaning to the word "decentralization." If miners (a much more centralized group than non-mining nodes) control all nodes, then the entire userbase is at the mercy of a group that can collude against them.

Ahh, I see, on a purely technical level. So if I run 5000 nodes, I become 1/2 the userbase & Bitcoin becomes x2 as "decentralized." GG.

No, you don't become half the userbase. But you might be a potential Sybil attacker. Give it a try, I say. Go ahead and test bitcoin's level of decentralization, and see if you can attack it.

Quote from: bargainbin on February 13, 2016, 03:19:03 PM

Quote

That the number of nodes, or a given proportion of nodes, can be obfuscated in the short term has nothing to do with that. It doesn't matter that you can't have a clear picture of the node software people are running, nor of their intentions....what matters is that without peers, bitcoin becomes a network based on trust.

But it already is! Gah.

Edit, TL;DR: Yes, Bitcoin would be more secure if every wallet was a full node. It would be more secure still if every wallet also hashed. This is not the case, nor is it likely to be. We rely on (trust) multiple middlemen, from BitPay to exchanges, and trust the people we transact with to actually send us something in return for the bits we send them. Number of non-mining nodes is irrelevant.

Regarding the bolded: no, it isn't. Not yet, anyway. As a bitcoin user, I don't rely on Bitpay, exchanges or any other third party -- that's only for fiat conversion, which is completely external to the bitcoin system. The faults of "trusting people we transact with" are due to "trusting people we transact with" -- not bitcoin. As I've said and provided evidence for several times, non-mining nodes are very much relevant to the decentralization and security of bitcoin.