Please run a full node

[kolloh]

Yeah, we need more miners rather than more full nodes I think to help make the network stronger. But miners require significant investments so most folks won't be setting up new miners even though its needed.

This is sigspam nonsense. There is more than enough mining hardware around. What isn't enough of in mining is enough spread of the hashrate. More people buying hardware won't fix that when the economies of scale and cheap electricity means most hardware is clustered in a small number of huge farms owned by very few people.

Yeah that is what I was implying. We need more new miners that aren't controlled by the same small group of people. New people buying mining hardware would indeed help spread the hashing power out which would help make the network stronger and less centralized into the power of these farms that are controlled by few people. However, as I was saying this would require a significant investment and likely wouldn't be profitable which is why there aren't many people willing to setup miners just to help the network grow stronger and less centralized.

[franky1]

its not just a mining game of who builds the biggest tower of blocks the fastest wins. where everyone then copies(like sheep) the fastest built tower..

=> but it is exactly like that.

(because there is no OTHER tower)

there are other towers.. you are just seeing it from the mid-endgame... you have not seen it from the start-midgame.

EG your only seeing the end result. you have not played out the scenarios from beginning to end

the reason there is now only one tower is because the diverse pools and nodes have learned their lessons. so you dont really see much drama of the different towers.
but now and again find out that they still have lessons to learn (orphans happen weekly)

[worldmobilecoin]

I will disagree.  More non-mining nodes doesn't necessarily make the network stronger if there are a sufficient number of nodes already.  Remember that only mining nodes can extend and secure the ledger. 

Agree with you. We need more 'real' mining nodes to make BTC more secure. Seems we have average 6-7K daily nodes. But mining still dominate by China even they only have 200-300 nodes.

[dinofelis]

Is there anything we can do to fight this monopoly? My understanding is that changing the structure of the Bitcoin network needs a hard fork, and a hard fork needs hashing power. Or we are just fu**ed? Is this the biggest flaw in Nakamoto's vision?

Well, the first thing one should realize is that in block chain land, *everything* is decided by an interaction between two classes of entities in the eco system:
- the "chain builders" on one hand
- the "users" on the other hand: the users are the people that are willing to OFFER VALUE FOR TOKENS.

In most cases, the chain builders get "rewards" for building the chain in the form of tokens on the chain they are building.  So in the end, the chain builders usually want to convert those tokens to value, by dumping them on the "users", who, for several reasons, desire to possess tokens on the chain at hand ; so to a certain extend, chain builders are sensitive to the willingness of users to pay for their chain.  On the other hand, users need the chain builders to be able to use the tokens they acquired, and to be able to sell them.

This is about the fundamental relationship between the entities in the crypto eco system.

So who's the boss ?  As we saw, both users and chain builders need one another: chain builders need users to buy their tokens ; users need chain builders to transact their tokens.  

As such:
a) users are the boss, because they are free not to buy chain builder's tokens
b) chain builders are the boss, because they can fix the rules by which users can access their possessions

So new buyers are the boss of the chain builders because they pay them, or not, and chain builders are the boss over stake holders because they give them access to their stake, or not.  But new buyers become stake holders.

As usual in such an eco-system, those that are more divided, are less powerful than those that are in agreement.

The most important point is however, this:
If chain builders are divided, then several chains can be built.   If chain builders are in agreement, then only one single chain is built.

If there is only one chain out there, and nobody else is building a chain, then the only option is to accept that chain as the ruling chain, or to leave the system (not buy any coins, and sell one's coins according to the requirements of the chain builders).  As such, chain builders in agreement have total technical control over all of the rules of the system, but they still need to seduce buyers.

So the only way to "attack" chain builders that are in agreement, apart from leaving the system, is:

a) to divide them or
b) to build other chains "oneself".

a) means: to give better terms to those going in your direction, than to what they are agreeing upon ; but most of the time, you would like them to have WORSE terms (lower fees, say) and less power.  So that's not going to work.

b) build another chain yourself remains.

This could be done in two ways:
a) change the PoW algorithm, making the miner's installations useless
b) change to a PoS style algorithm

In both cases, however, you are making a competing chain with the original one, and the question is: how will the users vote with their money ?  Users have equal amounts of coins on both chains, and will try to "dump" the chain of which they expect not to win, and will try to keep their coins on the chain that they expect to win in the market.  --> if you make a mistake, you lose all of your stake !

And this is where things get nasty.  How do you know that most users will prefer the new chain, and not stick to the old chain, or think that most users will stick to the old chain ?

After all, chances are that the old chain is considered "true bitcoin" (nothing changes !) and the new chain an alt coin.  It can even be that, if this is successful, did one open the flood gates, because what stops yet another entity from proposing YET ANOTHER "other chain" ?  Once you've successfully forked off, why not fork off N more times ?

It could work.  With ETH, it sort of worked.  The old chain, ETC, lived on, but an economic majority + devs went to the new chain (maybe also because the name was, remarkably, put on the new chain).  But ETH didn't have a "first mover brand name".

So, until one decides to divide the miners, or have the miners agree on a change, the option of making a totally new chain with another PoW or PoS, is going to be a most risky operation without any guarantee.

[franky1]

==> what "orphaning effect" ?  Orphaning occurs when OTHER MINERS decide to mine upon ANOTHER block than the orphaned block.  If miners decided to mine upon block A, with blocks B, C, D and E, there simply isn't any other set of blocks around that a P2P node could "prefer".  Suppose that a majority of P2P nodes decides to "orphan" block B.  But the miners have been building blocks C, D and E on top of B.  What happens ?

Well, these nodes stop.  They stop at block A.  And they can't find any other block that pleases them.  B wasn't according to their taste.  But there's NO OTHER BLOCK around that is built upon A.  Nobody has ever made a block on top of A with a higher PoW than the chain B,C,D,E.  In fact, nobody ever made a block on top of A.

but then pools see that all the merchants cant see their rewards for BCDE..
the users waiting on transactions cannot see BCDE

and pools have to wait until Z4 before they can even spend B...

so way way way before z4 (z4=going through the alphabet 4times(100 block confirm maturity)) occurs just to spend B.. pools realise they better make blocks according to rule A otherwise they have wasted half a day building b-zb-zb-zb-z that would be unspendable.

this is why instead of having orphans of 100 blocks deep where pools realise they cant spend their funds..
they only at most regularly have orphans of 1 block deep. because:
they know its foolish to keep mining b-z 4times.
they know its foolish to keep mining b-z 4times HOPING the delay would force 6000 nodes to download and waste a week re-syncing to new B rules.

pools only move to B rules if the nodes accept it, which pools find out about alot sooner than 6 hours+ when its time to spend.
pools can usually find out within a few minutes if their block is 'good' for the majority

some pools realise in 3 seconds

2017-01-29 06:59:12 Requesting block 000000000000000000cf208f521de0424677f7a87f2f278a1042f38d159565f5
2017-01-29 06:59:15 ERROR: AcceptBlock: bad-blk-length, size limits failed (code 16)

[dinofelis]

but now and again find out that they still have lessons to learn (orphans happen weekly)

These orphans have nothing to do with full nodes.  They have to do with the propagation delay of blocks amongst mining pools.
In other words, if, say, an orphaned block appears on average once or twice a week, say about every 400 blocks, it means that it takes on average 600 seconds / 400 blocks = 1.5 seconds for mined blocks to propagate to all miner pools
(rough estimation).

It means that within 1.5 seconds, a miner pool has received a valid block, has verified it, and starts mining a new block on top of it.  So one out of 400 times, a pool has found, by coincidence, a competing block, before realizing that a competitor was faster.

(BTW, this also indicates why miners are most probably directly connected: they could never have such a low orphaning rate if the 1MB blocks propagated through the P2P network: the delays between sending a block and having all other miners stop their work and start mining on top of yours would be too long, and would cause much more orphaning).

[franky1]

but now and again find out that they still have lessons to learn (orphans happen weekly)

These orphans have nothing to do with full nodes.  They have to do with the propagation delay of blocks amongst mining pools.
In other words, if, say, an orphaned block appears on average once or twice a week, say about every 400 blocks, it means that it takes on average 600 seconds / 400 blocks = 1.5 seconds for mined blocks to propagate to all miner pools
(rough estimation).

It means that within 1.5 seconds, a miner pool has received a valid block, has verified it, and starts mining a new block on top of it.  So one out of 400 times, a pool has found, by coincidence, a competing block, before realizing that a competitor was faster.

(BTW, this also indicates why miners are most probably directly connected: they could never have such a low orphaning rate if the 1MB blocks propagated through the P2P network: the delays between sending a block and having all other miners stop their work and start mining on top of yours would be too long, and would cause much more orphaning).

nooo
the MAJORITY of orphans happen now because of propogation delay....
 because pools are no longer risking changing the rules to make orphans happen for other reasons..

but if pools were to change the rules they would get orphaned

2017-01-29 06:59:12 Requesting block 000000000000000000cf208f521de0424677f7a87f2f278a1042f38d159565f5
2017-01-29 06:59:15 ERROR: AcceptBlock: bad-blk-length, size limits failed (code 16)

[Kprawn]

Is there anything we can do to fight this monopoly? My understanding is that changing the structure of the Bitcoin network needs a hard fork, and a hard fork needs hashing power. Or we are just fu**ed? Is this the biggest flaw in Nakamoto's vision?

No Satoshi actually envisioned that we would reach a stage where only big companies would be hosting nodes. We already see a scenario where only

people with fast internet and lots of disk space are capable of running FULL nodes. It takes a lot of bandwidth too, so you will have to take that into

consideration, if you want to run a FULL node. All of this can become pretty expensive, if you are living in some 3rd world country with expensive data

packages and slow internet.  

[Karpeles]

Most people will have only 8 connections or will shut down the node from time to time and will be syncing most of time, so they will be a burden for the network most of time.

Also I think the network is more than strong enough with the current nodes. It is more a future problem maybe than an urgent issue

[dinofelis]

==> what "orphaning effect" ?  Orphaning occurs when OTHER MINERS decide to mine upon ANOTHER block than the orphaned block.  If miners decided to mine upon block A, with blocks B, C, D and E, there simply isn't any other set of blocks around that a P2P node could "prefer".  Suppose that a majority of P2P nodes decides to "orphan" block B.  But the miners have been building blocks C, D and E on top of B.  What happens ?

Well, these nodes stop.  They stop at block A.  And they can't find any other block that pleases them.  B wasn't according to their taste.  But there's NO OTHER BLOCK around that is built upon A.  Nobody has ever made a block on top of A with a higher PoW than the chain B,C,D,E.  In fact, nobody ever made a block on top of A.

but then pools see that all the merchants cant see their rewards for BCDE..
the users waiting on transactions cannot see BCDE

Of course they can if they configure their wallet to connect directly to a miner pool node !   And do you think that a merchant or an exchange is going to stop its business for a day, while the competition accepts the new rules on the chain, and gets its node and wallet up and running again ?

Suppose that miners have announced that you should configure your node to 1.2 MB, because otherwise, you won't be able to see their chain.  Now, suppose that you need to send or receive 100 BTC for an important deal.  What are you going to do ?  Stop your node ?  Or accept their change ?  Suppose you are a merchant.  You simply want to be able to get bitcoins, and to send them for conversion.  You don't care about the technicalities, you simply want to use it.  What are you going to do ?  Connect to a stopped node, or connect to something that sees the current chain and accepts transactions ?

Suppose that you are an exchange.  Are you going to block withdrawals and deposits from customers, or are you going to update your node to suit the new chain ?  

As you say, the miners can wait to spend, they even have to wait to spend !  They just keep on mining the same chain, and you can chose to join them, and see your funds and be able to transact, or simply do nothing.  You upgrade, you are online again with bitcoin ; you don't upgrade, you are off the chain and off bitcoin.  What are you going to do ?

As a user, you are rather agnostic, and just want to use bitcoin.  As a full node owner, you can chose to have your node halted or to agree with the miner's protocol.  But by doing so, you are just going to get a user to ignore you, because the user wants a working chain to be able to transact on.

so way way way before z4 (z4=going through the alphabet 4times(100 block confirm maturity)) occurs just to spend B.. pools realise they better make blocks according to rule A otherwise they have wasted half a day building b-zb-zb-zb-z that would be unspendable.

Of course it will be spendable.  Nobody else can in any case spend what so ever during all of this time.  All those not accepting the chain are "taken hostage" too.  So who's going to give in first ?  The mining pools, that eventually would like to sell their coins, or the nodes that serve the users, that cannot do anything on the network any more until they accept the new chain ? 

Remember that most users don't have full nodes.  There are millions of users, and only a few thousand of full nodes.  Users have wallets that connect to a reference node - this could just as well be a miner node.

[franky1]

No Satoshi actually envisioned that we would reach a stage where only big companies would be hosting nodes. We already see a scenario where only

people with fast internet and lots of disk space are capable of running FULL nodes. It takes a lot of bandwidth too, so you will have to take that into

consideration, if you want to run a FULL node. All of this can become pretty expensive, if you are living in some 3rd world country with expensive data

packages and slow internet.  

1. his whole mindset (full context) is that there would be ~100k full nodes and no need for millions/billions of full nodes. the reason for this is:
a. the more nodes to propogate to makes it more delayed to get the data to everyone. so having 7billion full nodes actually hurts a decentralised network. (x degree's of separation)
b. not everyone NEEDS to be a full node
c. it would mainly be businesses and enthusiasts that NEED to run full nodes

[dinofelis]

but if pools were to change the rules they would get orphaned

2017-01-29 06:59:12 Requesting block 000000000000000000cf208f521de0424677f7a87f2f278a1042f38d159565f5
2017-01-29 06:59:15 ERROR: AcceptBlock: bad-blk-length, size limits failed (code 16)

No, if SOME pools would change the rules.  But we are considering the case where ALL pools fix the rules (the same ones, or the same change).

You are just printing what your local full node would tell YOU.  But if all miners are in agreement on the rules (old ones or new ones) - that's the case we consider - then there's no orphaning.  Because only miners can orphan blocks, by building on other blocks.

(btw, strictly speaking, an invalid block is not orphaned but rejected: the definition of orphaning is a VALID block that is not built upon ; by definition, blocks in the chain are valid).

[franky1]

Of course they can if they configure their wallet to connect directly to a miner pool node !   And do you think that a merchant or an exchange is going to stop its business for a day, while the competition accepts the new rules on the chain, and gets its node and wallet up and running again ?

your ignoring the past and creating scenario's that never happened..

pools learn in seconds if their block is going to be accepted.. they wont just continue for 16 hours HOPING merchants adapt..
pools learned this lesson the hard way years ago..

it seems you are trying too hard to downplay satoshi's consensus mechanism and trying too hard to make bitcoin sound like the fiat system.

its time you stop looking at the end result and thinking "its just a database" and instead actually look at all the different security mechanisms that make bitcoin completely different to fiat.

[dinofelis]

pools learn in seconds if their block is going to be accepted.. they wont just continue for 16 hours HOPING merchants adapt..
pools learned this lesson the hard way years ago..

.... by other miners.  We are in the scenario where all pools agree upon a protocol, and the full nodes want another one.  Pools only care about their block being accepted by other miners - if that's the case, their block is not orphaned, but is now part of the immutable history.  The few seconds is exactly that.

[dinofelis]

it seems you are trying too hard to downplay satoshi's consensus mechanism and trying too hard to make bitcoin sound like the fiat system.

No, on the contrary.  Satoshi introduced "consensus by majority of hash rate" to avoid any "majority of nodes".  Satoshi wrote that "only people wanting to make new coins should run full nodes", and he considered these to be "big data centers with special purpose hardware".

Now, Satoshi estimated that there would be less than 100 000, and he was right: there are only about 20.  20 is less than 100 000.

What Satoshi didn't realize or didn't want to say openly if he did, was that miners would pool together to increase their hash rate efficiency and that in a winner-takes all lottery, of course, pooling together is advantageous.  Instead of potentially winning one block every 6 months, you can be paid the same average amount by selling your hash power to a pool, delegating your decision on what block to mine.

  He didn't realize or he didn't want to say openly if he did, that economies of scale would bring the mining market into what every mature market does: a power distribution of market shares, concentrated in those areas of the world where the systems can work most efficiently.

In other words, an industry with about 20 entities sharing most of the market is normally to be expected from his design.

[franky1]

but if pools were to change the rules they would get orphaned

2017-01-29 06:59:12 Requesting block 000000000000000000cf208f521de0424677f7a87f2f278a1042f38d159565f5
2017-01-29 06:59:15 ERROR: AcceptBlock: bad-blk-length, size limits failed (code 16)

No, if SOME pools would change the rules.  But we are considering the case where ALL pools fix the rules (the same ones, or the same change).

You are just printing what your local full node would tell YOU.  But if all miners are in agreement on the rules (old ones or new ones) - that's the case we consider - then there's no orphaning.  Because only miners can orphan blocks, by building on other blocks.

and while say pools BCDE are creating their blocks ontop of B for 16 hours.

pool A gives nodes blocks that are rule A acceptable. and pool A get to spend the funds(pool Awins every 10 minutes, zero competition)

so 16 hours earlier.. it would be like

nodes 399,999 height rule A
poolb 400,000b height rule B
poola 400,000a height rule A

nodes 400,000a
poolb 400,001b
poola 400,001a

..16 hours later
nodes 400,100a
poolb 400,101b - dang it i cant spend 400,001 and it looks like i cannot spend the other 99 blocks either. dang it i wasted half a day
poola 400,101a - woo hoo i won every block for last 100, PARTY AT MY HOUSE 1250 btc to spend thanks B for being a dumb & orphaning urself for 16 hours

..16 hours later
nodes 400,101a
poolb 400,102a - ok i lost 100 rewards, ill just stick with rule A from now on.. i wont be changing the rules that easily again
poola 400,102a - dang it B learned his lesson.. ok guys party only every other block, we have competition again, seems they learned their lesson

..
which is why NOW when a pool sees their bloc getting ropped by the ntwork.. they wont continue for 16 hours they realise their mistake straight away.

(btw, strictly speaking, an invalid block is not orphaned but rejected: the definition of orphaning is a VALID block that is not built upon ; by definition, blocks in the chain are valid).

by the way strictly speaking an orphan was 'suppose' to be used in terms of, when a CHILD (newest addition) gets rejected purely because the parent(previous block) disappears..

but reality is a new block can be an orphan if the parent still exists but just gives up its child..

check the definition
An orphan is a child whose parents are dead or have permanently abandoned the child.

most pretend that "orphans" = parents are dead.. but the reality of human orphanages/foster care system is that children do not have to have dead parents to be classed as orphans.

[jonald_fyookball]

Daily reminder: Bitcoin is controlled by a pompous pimpled faced rich kid in China

says the idiot who recently exclaimed: "its my opinion Bitcoin is only at 10% capacity"  (when charts clearly show full blocks.)   

[dinofelis]

but if pools were to change the rules they would get orphaned

2017-01-29 06:59:12 Requesting block 000000000000000000cf208f521de0424677f7a87f2f278a1042f38d159565f5
2017-01-29 06:59:15 ERROR: AcceptBlock: bad-blk-length, size limits failed (code 16)

No, if SOME pools would change the rules.  But we are considering the case where ALL pools fix the rules (the same ones, or the same change).

You are just printing what your local full node would tell YOU.  But if all miners are in agreement on the rules (old ones or new ones) - that's the case we consider - then there's no orphaning.  Because only miners can orphan blocks, by building on other blocks.

(btw, strictly speaking, an invalid block is not orphaned but rejected: the definition of orphaning is a VALID block that is not built upon ; by definition, blocks in the chain are valid).

and while say pools BCDE are creating their blocks ontop of B for 16 hours.

pool A gives nodes blocks that are rule A acceptable. and pool A get to spend the funds(pool Awins every 10 minutes, zero competition)

First of all, no. If you want to prove that full nodes can impose their verification onto mining pools, you must be able to prove that in the case where all mining pools are agreeing amongst themselves ; otherwise you do not distinguish the effects of a hard fork, and the effects of the power of the full nodes, which is what you want to prove.

In other words, if you want to prove that full non mining nodes have decision power over miners, you must be able to prove that they can impose their rules also over a set of mutually agreeing mining pools.  

So the Gedanken experiment is simply that: all mining pools (of any significance, say the 20 biggest) AGREE on a set of rules, and the full nodes DISAGREE.  It is in this Gedanken Experiment that you have to show how the full nodes are going to impose their rule set over those of the mutually agreeing miners.

We are NOT talking about users, and not talking about miners making different chains and forking.  We are talking about how full nodes "keep the miner pools in check".  So show me how the large majority of full nodes is going to impose its rules against all miner pools mutually agreeing.

Moreover, even outside of the argument, the forking pool that agrees with the full nodes will not make a block every 10 minutes: it will make a block at exactly the same rate as it was winning blocks when in competition, because that's given by the constant difficulty.  So if pool A was winning a block every 3 hours, it is still going to be able to make a block every 3 hours, grossly.  And, again, this doesn't prove anything about the power of full nodes ; this only indicates what happens when pools fork.  Stick to the case where all miners agree on a set of rules, where the full nodes don't agree with, if you want to show the power of full nodes imposing their rules on miners.

[franky1]

We are in the scenario where all pools agree upon a protocol, and the full nodes want another one.  Pools only care about their block being accepted by other miners

nope. because IF pools right now made blocks 465623-465723 that were, say using MD5 hashing..

by this evening.. they would find out that all the merchants wont see their rewards of 465622+
the pools would have most definetly realised that merchants wont buy their 465622 reward by at most 465623

Stick to the case where all miners agree on a set of rules, where the full nodes don't agree with, if you want to show the power of full nodes imposing their rules on miners.

so pools wont continue making md5 hashed blocks for 16 hours because 1 pool will see a nice easy income by switching back to sha256 and win every block that is spendable to the merchants and have no competition

[dinofelis]

We are in the scenario where all pools agree upon a protocol, and the full nodes want another one.  Pools only care about their block being accepted by other miners

nope. because if pools right now made blocks 465623-465723 that were, say using MD5 hashing..

by this evening.. they would find out that all the merchants wont see their rewards of 465622
the pools would have most definetly realised that merchants wont buy their 465622 reward by at most 465623

Well, they would realize that all merchants are simply locked out of all transactions, unless merchants use the new rule software on their full node, or connect their wallet to a miner node.  Because there aren't any other transactions in accepted blocks.

So yes, miners can't cash out their coins to users, as long as ALL USERS decide not to upgrade, and accept being locked out of bitcoin, their funds and their transactions at the same time.  The question is why users would keep themselves locked out, while they can be up and running with bitcoin by simply running nodes that accept the existing block chain, or by connecting their wallets to such nodes, of which the miner pool nodes are of course examples.