PoW vs PoS conundrum - presenting a new form of PoA.

[100KN]

Hello Everyone,

I present to you a 'premise' and a 'proposal' for a better future cryptocurrency, I have spent a few days formulating this in my mind and then on paper. Its a long post and fairly technical, but spend a few minutes following the logic trail and let me know what you think. Criticisms most welcome!


Proof of Work (Bitcoin)

• - Energy intensive. Not ecological.
  - Centralization. Unfair wealth distribution.
  - Unsustainable, may not scale.
  - Transaction fees not mandatory
  - Reducing block rewards
  - Commercial Mining
  - 51% attack threat

Proof of Stake (Eth/Nxt,NEM):

• - Not trustless
  - No verifiable longest chain.
  - No 'work' involved
  - Blockchain can fork
  - No single source of truth, history can be re-written
  - Checkpoints need trust.
  - Economic inequality, rich get richer.
  - Centralization, attack, total collapse.

What I've gathered so far:

• - Work (wastage?) is inevitable.
  - Transaction fees are necessary.
  - Mining is essential economically.
  - Monetarism - controlled inflation
  - 100% premined will not work eventually.
  - Fungibility is essential, but so is transparency. Ideally should offer flexibility, like ShadowCash does; Monero cannot be mainstream.

Proposing a modified "Proof of Activity" (hybridized PoW/PoS)

Gist:
First shortlist lucky miners to sign next block based on stake (even distribution; weighted on stake ); next offer them a 'work' i.e. a complex math problem, (ASIC 'resistant', not asic proof), the first one to solve gets to sign the block and gets the reward.
In reality there is no prior shortlisting and handing them a math problem.
Each node is able to quickly self verify if it qualifies for the stake based lottery**( explained below) and if 'yes', it can begin solving the math problem**(explained below).

------------------------------------------------------------------

Stake based selection:

Design Considerations:

• 1 - Give weight-age to stake/investment.
  2 - Reward 'activity', securing the network.
  3 - Balance 1 & 2 such that Miners will NOT consider diversifying investment into accounts.
  4 - Prevent centralization to possible extent.

Algorithm:

• L = Chain window length , e.g. 50
  N = circular counter (1-50), changes with each block
  BhX= Hash of Block N, in base-62, 64 digit long
  BhY = Hash of Block (L-N), in base-62, 64 digit long
  A = node's address in base 62.

Each node performs,

• M (mash) =  Fa(BhX,BhY), number in base-62, 64 digit long.  i.e. Add the two base-62 BhX & BhY, apply modulo '%' to limit length to 64.
  Ss (score)= Fcb(M,A) , see link for the basic game, https://www.rosettacode.org/wiki/Bulls_and_cows , but extended as below:
  - alphanumeric matching in same position will gather 3 points.
  - alphabet matching in same position but different cases (upper,lower) will gather 2 points.
  - alphanumeric matching in different position will gather 1 point.

• V = UTXOs of node's address 'A' in currency units.
  
  Dfs = Difficulty factor(stake), range 0.0 to 1.0
  
  Q = Dfs * 62 - (log (V) / log (√2))
  

Note:

• - So you can interpret Q value as number of character matches a node is able to find.
  - Scroll to the end to find a sample tabulation of V & Q values, for Dfs of 1.0 (unlikely to be used)

A match can gather either of 1,2 or 3 points. So on average a match accumulates 2 points.

if (Ss > 2*Q) then mine();

------------------------------------------------------------------

Proof of Work:

Design Considerations:

• 1 - Common everyday computers / miners should be able to participate.
  2 - Do not use existing ASIC friendly algorithms
  3 - Solving should be limited to heuristic techniques at best.
  4 - The obvious difficult to solve, but easy to verify.
  5 - Invent one, because its fun
  

Algorithm:

• Imagine an everyday number lock, only this one has 62 wheels instead of 3.
  Each wheel/dial has 62 unique digits/characters, A-Za-z0-9.
  Each wheel is NOT independent, so turning each wheel affects other wheels.
  Moving a (any) wheel by 1 step up, moves all the following (progressing to the right) wheels in different number of steps, following some (Fibonacci?) pattern ,and also alternating Up and Down. It also wraps around until it reaches the neighbor on the left.
  Also consider the wheels can be turned on Y axis too, and they have similar effects of moving their neighbors below them alternating Left and Right.
  Now, the node is supposed to come up with a short route to align numbers on the pseudo 3 dimensional wheel such that,
  

• Sw = Score for work done.
  Dw = Difficulty for work/mining
  CoR = values of a single column or row that is potentially matching to the address A, based on the scoring system described above (Bulls and Cows based)

• Sw = Fcb(CoR,A)
  
  If Sw > Dw  sign next block and claim block reward.

Of course this should be easy to verify by the remaining nodes as the winning miner will publish the path(Pt) to align the dials which contain the interested CoR,
The path could be an unoptimized string in this format:

• N = Column/Row number
  Di = Up/Down
  St = Steps to move, min 1 to max 10
  

[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St]

• Pt = Path to align the dials, limited to 16 instructions.
  pCoR = pointer to CoR, contains the column number or row number of the winning combination.

***sample tabulation of V & Q values, for Dfs of 1.0.
 

• 62    10
  61    14
  60    20
  59    28
  58    40
  57    56
  56    79
  55    111
  54    156
  53    220
  52    311
  51    438
  50    617
  49    871
  48    1,228
  47    1,731
  46    2,441
  45    3,441
  44    4,852
  43    6,842
  42    9,647
  41    13,602
  40    19,179
  39    27,042
  38    38,129
  37    53,762
  36    75,805
  35    106,885
  34    150,708
  33    212,498
  32    299,622
  31    422,467
  30    595,678
  29    839,906
  28    1,184,268
  27    1,669,817
  26    2,354,442
  25    3,319,764
  24    4,680,867
  23    6,600,022
  22    9,306,031
  21    13,121,504
  20    18,501,321
  19    26,086,862
  18    36,782,476
  17    51,863,291
  16    73,127,240
  15    103,109,408
  14    145,384,266
  13    204,991,815
  12    289,038,459
  11    407,544,227
  10    574,637,360
  9    810,238,677
  8    1,142,436,535

[1714865199]

1714865199

[1714865199]

1714865199

[1714865199]

1714865199

[1714865199]

1714865199

[densuj]

Hello Everyone,

I present to you a 'premise' and a 'proposal' for a better future cryptocurrency, I have spent a few days formulating this in my mind and then on paper. Its a long post and fairly technical, but spend a few minutes following the logic trail and let me know what you think. Criticisms most welcome!


Proof of Work (Bitcoin)

• - Energy intensive. Not ecological.
  - Centralization. Unfair wealth distribution.
  - Unsustainable, may not scale.
  - Transaction fees not mandatory
  - Reducing block rewards
  - Commercial Mining
  - 51% attack threat

Proof of Stake (Eth/Nxt,NEM):

• - Not trustless
  - No verifiable longest chain.
  - No 'work' involved
  - Blockchain can fork
  - No single source of truth, history can be re-written
  - Checkpoints need trust.
  - Economic inequality, rich get richer.
  - Centralization, attack, total collapse.

What I've gathered so far:

• - Work (wastage?) is inevitable.
  - Transaction fees are necessary.
  - Mining is essential economically.
  - Monetarism - controlled inflation
  - 100% premined will not work eventually.
  - Fungibility is essential, but so is transparency. Ideally should offer flexibility, like ShadowCash does; Monero cannot be mainstream.

Proposing a modified "Proof of Activity" (hybridized PoW/PoS)

Gist:
First shortlist lucky miners to sign next block based on stake (even distribution; weighted on stake ); next offer them a 'work' i.e. a complex math problem, (ASIC 'resistant', not asic proof), the first one to solve gets to sign the block and gets the reward.
In reality there is no prior shortlisting and handing them a math problem.
Each node is able to quickly self verify if it qualifies for the stake based lottery**( explained below) and if 'yes', it can begin solving the math problem**(explained below).

------------------------------------------------------------------

Stake based selection:

Design Considerations:

• 1 - Give weight-age to stake/investment.
  2 - Reward 'activity', securing the network.
  3 - Balance 1 & 2 such that Miners will NOT consider diversifying investment into accounts.
  4 - Prevent centralization to possible extent.

Algorithm:

• L = Chain window length , e.g. 50
  N = circular counter (1-50), changes with each block
  BhX= Hash of Block N, in base-62, 64 digit long
  BhY = Hash of Block (L-N), in base-62, 64 digit long
  A = node's address in base 62.

Each node performs,

• M (mash) =  Fa(BhX,BhY), number in base-62, 64 digit long.  i.e. Add the two base-62 BhX & BhY, apply modulo '%' to limit length to 64.
  Ss (score)= Fcb(M,A) , see link for the basic game, https://www.rosettacode.org/wiki/Bulls_and_cows , but extended as below:
  - alphanumeric matching in same position will gather 3 points.
  - alphabet matching in same position but different cases (upper,lower) will gather 2 points.
  - alphanumeric matching in different position will gather 1 point.

• V = UTXOs of node's address 'A' in currency units.
  
  Dfs = Difficulty factor(stake), range 0.0 to 1.0
  
  Q = Dfs * 62 - (log (V) / log (√2))
  

Note:

• - So you can interpret Q value as number of character matches a node is able to find.
  - Scroll to the end to find a sample tabulation of V & Q values, for Dfs of 1.0 (unlikely to be used)

A match can gather either of 1,2 or 3 points. So on average a match accumulates 2 points.

if (Ss > 2*Q) then mine();

------------------------------------------------------------------

Proof of Work:

Design Considerations:

• 1 - Common everyday computers / miners should be able to participate.
  2 - Do not use existing ASIC friendly algorithms
  3 - Solving should be limited to heuristic techniques at best.
  4 - The obvious difficult to solve, but easy to verify.
  5 - Invent one, because its fun
  

Algorithm:

• Imagine an everyday number lock, only this one has 62 wheels instead of 3.
  Each wheel/dial has 62 unique digits/characters, A-Za-z0-9.
  Each wheel is NOT independent, so turning each wheel affects other wheels.
  Moving a (any) wheel by 1 step up, moves all the following (progressing to the right) wheels in different number of steps, following some (Fibonacci?) pattern ,and also alternating Up and Down. It also wraps around until it reaches the neighbor on the left.
  Also consider the wheels can be turned on Y axis too, and they have similar effects of moving their neighbors below them alternating Left and Right.
  Now, the node is supposed to come up with a short route to align numbers on the pseudo 3 dimensional wheel such that,
  

• Sw = Score for work done.
  Dw = Difficulty for work/mining
  CoR = values of a single column or row that is potentially matching to the address A, based on the scoring system described above (Bulls and Cows based)

• Sw = Fcb(CoR,A)
  
  If Sw > Dw  sign next block and claim block reward.

Of course this should be easy to verify by the remaining nodes as the winning miner will publish the path(Pt) to align the dials which contain the interested CoR,
The path could be an unoptimized string in this format:

• N = Column/Row number
  Di = Up/Down
  St = Steps to move, min 1 to max 10
  

[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St],[N][Di][St]

• Pt = Path to align the dials, limited to 16 instructions.
  pCoR = pointer to CoR, contains the column number or row number of the winning combination.

***sample tabulation of V & Q values, for Dfs of 1.0.
 

• 62    10
  61    14
  60    20
  59    28
  58    40
  57    56
  56    79
  55    111
  54    156
  53    220
  52    311
  51    438
  50    617
  49    871
  48    1,228
  47    1,731
  46    2,441
  45    3,441
  44    4,852
  43    6,842
  42    9,647
  41    13,602
  40    19,179
  39    27,042
  38    38,129
  37    53,762
  36    75,805
  35    106,885
  34    150,708
  33    212,498
  32    299,622
  31    422,467
  30    595,678
  29    839,906
  28    1,184,268
  27    1,669,817
  26    2,354,442
  25    3,319,764
  24    4,680,867
  23    6,600,022
  22    9,306,031
  21    13,121,504
  20    18,501,321
  19    26,086,862
  18    36,782,476
  17    51,863,291
  16    73,127,240
  15    103,109,408
  14    145,384,266
  13    204,991,815
  12    289,038,459
  11    407,544,227
  10    574,637,360
  9    810,238,677
  8    1,142,436,535

It is nice proposal and ideas, unfortunately it is suitable for developers of coins
but the problem are there the developers whom accept your proposal?
It is need much discussion about your ideas include weakness of your theory.
Or I think you must make altcoins that suitable with your ideas and proposal
i don't know about that and i am not developers of coin.

Let's wait until there are developers or other users whom has knowledge
building altcoins and I hope it is not becoming scam coins.

[dinofelis]

What your proposal actually does, is to reduce the "waste" by limiting the competition.  However, purely monetary-wise, the "waste" of PoW is necessary to KILL the seigniorage.  The very idea that someone could EARN more than a competitive margin (determined by the market) by minting coins, and hence profit from seigniorage, is always problematic with a monetary system.  The bitcoin-like PoW kills all that, up to a competitive margin, which is considered the reward for maintaining the network running.  The fact that bitcoin's PoW system is wasteful, is monetary-wise a necessity.  All new value created by the minting of coins should be destroyed.

If you do not do that, you get the PoS paradigm, where new coins created will give their almost net gain to the one allowed to mint them.  The problem with any PoS mechanism is that you get the compound interest formula in one way or another: the more you have, the more you will get.  If you "top off" the reward, you will incite big stake holders to make sybil accounts.  And if you do not top off the reward, you will get the compound interest formula.

Now, the compound interest formula diverges.  Take just any randomly distributed set of stakes, and apply long enough the compound interest formula to it, and you'll end up having one account reaching more than 90%, more than 99%, more than 99.99% ... of the total amount of coins in the system.  That can take a long time, of course depending on the "interest rate".  But it will end up centralizing all coins into one entity.

In fact, PoW diverges in the same way, but much, much slower, because the "interest" in PoW is the margin, which is determined by economies of scale, and at least, most of the "benefit" got wasted, so nobody got immensely rich with it.

With PoS you get a rich king ; with PoW you get a gardian of the biggest rubble pile.
If you combine both, you'll get a kind of rich king sitting on a rubble pile.

[kiklo]

PoW is a failure, energy wise and Centralization caused by economics,
their are no ways to solve it , it is an evolutionary dead end.

PoS , the longest chain with the most difficulty does win,
I don't know why so many people can't grasp that.

Nothing at stake is a load of bullshit , by G. Maxwell .
What else do you expect from a BTC/Monero/LN Supporter.
An imaginary problem, that is a non issue.
https://bitcointalk.org/index.php?topic=1709776.0

PoS as the consensus system is perfect,
the problem comes when you try to make a consensus system into a Interest Bearing Savings Account.

The only solution is to switch PoS to an ultra low inflation rate. 
Which everyone complains won't work, because their is no reason for stakers to defend the network,
to which I will reply , the stakers will have a greater reason to stake even an ultra low inflation PoS coin, than BTC Full Node operators currently have. 

 

[100KN]

I am indeed trying to reduce the waste by selectively allowing only a subset of miners to compete for the next block. I am aware of the divergence behavior of POS, I was trying to balance the 'Stake' to 'Work' and I think divergence can be controlled to a reasonable degree.

1 - Give weight-age to stake/investment.
2 - Reward 'activity', securing the network.
3 - Balance 1 & 2 such that Miners will NOT consider diversifying investment into accounts.
4 - Prevent centralization to possible extent.

In the PoA that I proposed,  lets say there are 1000 stake holders in the network, the selection algorithm selects only say 333 stakeholders for the mining effort to sign the next block. The remaining 666 are idling.
We have no choice but to offer relatively easier 'work' to miners with bigger account balances. If we don't give weightage(does not have to be linear) to their balances, they would diversify their account balance in multiple mining rigs (farms) will all churn away useful resources. Instead they can now sit back and enjoy the dividends from a single miner.

Maybe I am missing something here but,

• - The concern of having offered easier work to the rich is not something that needs to be solved. Everyone's money is working as hard as anybody elses. How can we selectively target one's(rich) money to perform less efficiently in earning dividends?
   - If one rich miner has 50% of the entire wealth, and the rest share the remaining 50%, if after a few years of PoA, if the wealth distribution remains the same 50%-50%, is it still a concern?
   - Are you saying that you'd rather have the rich burn money/resources by mining (securing the network) to safeguard their prior investment, thus diminishing their wealth?

What are your thoughts on Block rewards & transaction fees? How should they be ideally?
I am trying to find ways to destroy money as well. Is this sensible?

Its challenging to solve this in a economically stable yet, remain secure (long-range,sybil, forking) and ecological.

[dinofelis]

PoW is a failure, energy wise and Centralization caused by economics,
their are no ways to solve it , it is an evolutionary dead end.

PoS , the longest chain with the most difficulty does win,
I don't know why so many people can't grasp that.

In pure PoS, there is no difficulty.  What you are talking about is a mixture of PoS and PoW, like the OP presents, that is, there is a PoS mechanism that *allows a limited number of stake holders* to compete PoW-wise for a block.

PoS as the consensus system is perfect,
the problem comes when you try to make a consensus system into a Interest Bearing Savings Account.

Indeed.

The only solution is to switch PoS to an ultra low inflation rate. 

Well, the "inflation rate" is the coin creation rate in the end.  No matter what emission scheme you use, you will get a significant inflation rate in the beginning, or you have to start with a pre-mine.

What a restrictive PoW scheme does (like the OP presents, and like you do, even though you call it PoS) is to avoid full competition for the seigniorage, so that not most of the seigniorage is burned.  By restricting the number of competitors for each block, you avoid that they enter into competition to the level of where all the seigniorage is wasted, so that stake holders can pocket in the remaining seigniorage.  In order to reduce the divergence of this "interest bearing savings account" which you nevertheless create, you propose a very low inflation rate.  However, that very low inflation rate could be applied to a pure PoW system too.  The only thing you trade off is "interest bearing savings account" (that HAS to absorb all the non-burned seigniorage) versus "wasting more electricity and burning all seigniorage".

In other words, in as much as you render less efficient the centralisation due to economies of scale for the second aspect, you render more efficient the divergence of the compound interest formula of the interest bearing savings account.

You have a "quantity of seigniorage" to handle (namely, ALL the coins apart from a premine), and you can divide this in an amount of "interest bearing" and an amount of "burning PoW", but both mechanisms have in them, a divergence towards centralisation: the first one is obvious (compound interest), the second one is more involved (economies of scale).

If you want less of the second, you'll get more of the first.  And "lowering the inflation rate" only stretches this thing in time.

I'm sure there can be optimisations, to find the right combination between "value appreciation of the market cap of the coin", "inflation rate", and the cursor between "interest" and "PoW", to push the point of centralisation as far as possible in the future, but it will happen.

At least, the PoW scheme has the advantage of wasting almost all seigniorage and securing maximally the network.

Which everyone complains won't work, because their is no reason for stakers to defend the network,
to which I will reply , the stakers will have a greater reason to stake even an ultra low inflation PoS coin, than BTC Full Node operators currently have. 

But bitcoin full node operators have absolutely no "weight" or importance in the security of the bitcoin network (apart fluidifying a bit more the transmission of data).  It is a big misunderstanding that bitcoin full nodes contribute to any form of network security.  If they don't mine, they aren't worth anything.  In the bitcoin network, only the miners and the users matter.  The miners for the chain, and the users for the market cap.  You could take out all non-mining nodes, and that wouldn't change much (apart for some latencies in the network when transmitting transactions).

[dinofelis]

I am indeed trying to reduce the waste

My idea is that purely monetary wise, this waste is essential.  The economies of scale that lead to centralisation are a problem, but the waste is needed to kill off the seigniorage.

[thebatletbet]

iam coin user interest , previllage, like this POS coin
because us POS coin only stake in wallet can return coin
if pow must mining and high quality hardware can get coin

[shinratensei_]

PoW is a failure, energy wise and Centralization caused by economics,
their are no ways to solve it , it is an evolutionary dead end.

I just realize if that's like the critical criticism for bitcoin. 

[kiklo]

PoW is a failure, energy wise and Centralization caused by economics,
their are no ways to solve it , it is an evolutionary dead end.

PoS , the longest chain with the most difficulty does win,
I don't know why so many people can't grasp that.

In pure PoS, there is no difficulty.  What you are talking about is a mixture of PoS and PoW, like the OP presents, that is, there is a PoS mechanism that *allows a limited number of stake holders* to compete PoW-wise for a block.

Target is the BlockSpeed.

PoW Difficulty
Difficulty is a measure of how difficult it is to find a hash below a given target.

PoS Difficulty
Difficulty is a measure of how difficult it is to Stake a Block of Coins below a given target.
https://github.com/novacoin-project/novacoin/wiki/Proof-of-stake

PoS as the consensus system is perfect,
the problem comes when you try to make a consensus system into a Interest Bearing Savings Account.

Indeed.

The only solution is to switch PoS to an ultra low inflation rate.  

Well, the "inflation rate" is the coin creation rate in the end.  No matter what emission scheme you use, you will get a significant inflation rate in the beginning, or you have to start with a pre-mine.

What a restrictive PoW scheme does (like the OP presents, and like you do, even though you call it PoS) is to avoid full competition for the seigniorage, so that not most of the seigniorage is burned.  By restricting the number of competitors for each block, you avoid that they enter into competition to the level of where all the seigniorage is wasted, so that stake holders can pocket in the remaining seigniorage.  In order to reduce the divergence of this "interest bearing savings account" which you nevertheless create, you propose a very low inflation rate.  However, that very low inflation rate could be applied to a pure PoW system too.  The only thing you trade off is "interest bearing savings account" (that HAS to absorb all the non-burned seigniorage) versus "wasting more electricity and burning all seigniorage".

In other words, in as much as you render less efficient the centralisation due to economies of scale for the second aspect, you render more efficient the divergence of the compound interest formula of the interest bearing savings account.

You have a "quantity of seigniorage" to handle (namely, ALL the coins apart from a premine), and you can divide this in an amount of "interest bearing" and an amount of "burning PoW", but both mechanisms have in them, a divergence towards centralisation: the first one is obvious (compound interest), the second one is more involved (economies of scale).

If you want less of the second, you'll get more of the first.  And "lowering the inflation rate" only stretches this thing in time.

I'm sure there can be optimisations, to find the right combination between "value appreciation of the market cap of the coin", "inflation rate", and the cursor between "interest" and "PoW", to push the point of centralisation as far as possible in the future, but it will happen.

At least, the PoW scheme has the advantage of wasting almost all seigniorage and securing maximally the network.

Could PoW be done with an extremly low inflations, sure no dispute there.
However due to the economics involved it won't be, PoW mining energy & ASICS requirements costs are constantly growing for a successful PoW coin, because it is PoW it will become unaffordable.
PoS does not have the economic flaws of an ever increasing energy price or ASICS purchases, which is why it is the only method that will currently work.

Which everyone complains won't work, because their is no reason for stakers to defend the network,
to which I will reply , the stakers will have a greater reason to stake even an ultra low inflation PoS coin, than BTC Full Node operators currently have.  

But bitcoin full node operators have absolutely no "weight" or importance in the security of the bitcoin network (apart fluidifying a bit more the transmission of data).  It is a big misunderstanding that bitcoin full nodes contribute to any form of network security.  If they don't mine, they aren't worth anything.  In the bitcoin network, only the miners and the users matter.  The miners for the chain, and the users for the market cap.  You could take out all non-mining nodes, and that wouldn't change much (apart for some latencies in the network when transmitting transactions).

Hmm, my understanding was ASICS only process the PoW calculations, They don't maintain Blockchain ,
Only Full Nodes maintain the blockchain where others can download it, without the blockchain you have no transaction history.
No transaction history no BTC.  
http://www.coindesk.com/bitcoin-nodes-need/

For example, in order to validate and relay transactions, bitcoin requires more than a network of miners processing transactions, it must broadcast messages across a network using 'nodes'.
This is the first step in the transaction process that results in a block confirmation.

To function to its full potential, the bitcoin network must not only provide an avenue for transactions, but also remain secure.
By using a number of randomly selected nodes, the network can reduce the problem of double spending – when a user attempts to spend the same digital token twice.

However, bitcoin doesn't just need nodes, it requires lots of fully functioning nodes –
nodes that have the bitcoin core client on a machine instance with the complete block chain.
The more nodes there are, the more secure the network is.

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

What makes a full node?

Full nodes download every block and transaction and check them against Bitcoin's core consensus rules.
Here are examples of consensus rules, though there are many more:

    Blocks may only create a certain number of bitcoins. (Currently 12.5 BTC per block.)
    Transactions must have correct signatures for the bitcoins being spent.
    Transactions/blocks must be in the correct data format.
    Within a single block chain, a transaction output cannot be double-spent.

If a transaction or block violates the consensus rules, then it is absolutely rejected, even if every other node on the network thinks that it is valid.
This is one of the most important characteristics of full nodes: they do what's right no matter what.
For full nodes, miners actually have fairly limited power: they can only reorder or remove transactions, and only by expending a lot of computing power.
A powerful miner is able to execute some serious attacks, but because full nodes rely on miners only for a few things, miners could not completely change or destroy Bitcoin.

 

[dinofelis]

When Satoshi considered full nodes, he considered them to be mining too.

However, bitcoin doesn't just need nodes, it requires lots of fully functioning nodes –
nodes that have the bitcoin core client on a machine instance with the complete block chain.
The more nodes there are, the more secure the network is.[/color]

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

What makes a full node?

Full nodes download every block and transaction and check them against Bitcoin's core consensus rules.
Here are examples of consensus rules, though there are many more:

    Blocks may only create a certain number of bitcoins. (Currently 12.5 BTC per block.)
    Transactions must have correct signatures for the bitcoins being spent.
    Transactions/blocks must be in the correct data format.
    Within a single block chain, a transaction output cannot be double-spent.

If a transaction or block violates the consensus rules, then it is absolutely rejected, even if every other node on the network thinks that it is valid.
This is one of the most important characteristics of full nodes: they do what's right no matter what.
For full nodes, miners actually have fairly limited power: they can only reorder or remove transactions, and only by expending a lot of computing power.
A powerful miner is able to execute some serious attacks, but because full nodes rely on miners only for a few things, miners could not completely change or destroy Bitcoin.

Ask yourself what happens if:

1) all non-mining full nodes are in disagreement with a block
2) 70% of all mining nodes are in agreement with that block.

Answer: the mining nodes decide, and whether the non-mining full nodes reject those blocks or not, doesn't matter.  You, as a user, can decide (because you run a full node, or because you contact a full node) that the currently used block chain is not in agreement with the bitcoin rules, but that doesn't matter, the miners decide and make the chain.  If full nodes reject that chain, this doesn't matter.  They will just stop updating their local version of the chain, but no "good" chain will appear out of nothing.

In other words, full non-mining nodes can be aware that the rules aren't followed (that they have changed, in other words, that bitcoin has forked, in other words) but apart from noticing that, they can't do anything about it.

Users can still send their transactions directly to the mining nodes, if the non-mining nodes don't agree.

[kiklo]

Ask yourself what happens if:

1) all non-mining full nodes are in disagreement with a block
2) 70% of all mining nodes are in agreement with that block.

It depends,
if 70% of the mining ASICS agree the block award is 36 BTC per block, then the full nodes will block its entry into the blockchain.
Mining ASICS will fork off and Full Nodes will protect the BlockChain from corruption.  

Or if the Full Nodes detect that the coins have already been spent, it will deny its entry into the blockchain, (Preventing a double spend.)

if 70% of the mining ASICS agree and the block reward is 12.5 BTC, then the Full Node finds it meets all configured criteria , then it will allow its entry into the blockchain.

Miners Power is to control which Transactions are in a block , Full Nodes Prevent them from cheating the reward systems and easily counterfeiting coins with double spends.


 

FYI:
https://en.bitcoin.it/wiki/Full_node

If a transaction or block violates the consensus rules, then it is absolutely rejected, even if every other node on the network thinks that it is valid.
This is one of the most important characteristics of full nodes: they do what's right no matter what.

FYI2:  I think you should read the following. You confusing a node with an asic.

an ASIC miner has NO hard drive.
an ASIC miner does NOT store the blockchain
an ASIC miner is NOT affected by the blocksize in any way.
an ASIC miner just receives a hash.. (small string of text) and it processes that string of text using a puzzle to create another string of text

it does not matter how much data a block has, the hash an ASIC miner receives stays the same length.

do not blame the block size on miners!!

blame it on PEOPLE thinking that USERS NODES cannot cope. when in reality USER NODES can cope
blame it on PEOPLE stupidly thinking the block bloat needs to be gigabytes a day and terrabytes a month.

blame it on PEOPLE who ignore the rational consensus since last year that is 2-4mb a block .. no 80mb-100gb(fake irrational doomsdays)
yes that is right for the last year the compromise has been 2-4mb.. so will people stop the irrational random numbers needed tomorrow..

FYI3:
In Proof of Stake , All Nodes that can stake are Full Nodes.
Same Validation principle as PoW Full Nodes, say a Staker modifies his individual software to produce 3X the amount of coins per block,
he will generate a Block , but all of the other PoS Wallets will deny his block entry into the blockchain, because it fails group consensus.  
He will only generate Orphans all day.

[dinofelis]

Ask yourself what happens if:

1) all non-mining full nodes are in disagreement with a block
2) 70% of all mining nodes are in agreement with that block.

It depends,
if 70% of the mining ASICS agree the block award is 36 BTC per block, then the full nodes will block its entry into the blockchain.
Mining ASICS will fork off and Full Nodes will protect the BlockChain from corruption.  

And how would they do that ?  They will *observe* corruption, but what can they do about it ?  They cannot add any block.  They cannot make any forked chain (if they do, then they become mining nodes, which we supposed they weren't).

What does a non-mining, full node actually do, apart from verifying FOR HIMSELF whether the block chain that is distributed elsewhere actually corresponds to his own standards and wishes, or not ?  Once you see that the block chain that you downloaded from peers, doesn't correspond to what you thought it had to be, you can only refuse to use it, but that's about it.

Of course, when I talk about miners, I'm talking about the FULL NODES that mine (whether they are individual miners, or pools).  I'm not talking about JUST the mining hardware. I'm talking about the node that takes the decision on what block chain to build the next block.

Let us do a gedanken experiment.  Again, in what follows, the what I call a miner is the mining node, that decides on what block to work, on top of what chain.  THOSE are the (only) nodes that matter, as I will show.  A non-mining node (as I have one at home) doesn't have the slightest bit of "consensus power".

A fundamental modification of the protocol is proposed (like you say, the 36 BTC per block reward).  

We consider this:

EDIT: oops, I messed up, which made this incomprehensible.  Options corrected.

1) most miners are against it, but most non mining full nodes want to apply it, and users don't want want it.

2) most miners are against it, but most non mining full nodes want to apply it, and users want it.

3) most miners are in favor of it, but most non mining full nodes don't want it, and users want it.

4) most miners are in favor of it, but most non mining full nodes don't want it, and users don't want it.


What happens ?

In case 1, miners will still continue to mine normal block awards.  Most non-mining full nodes, who updated their code, will reject the block chain miners produce.  As users want it this way, they will short-circuit the non-mining nodes, and send their transactions directly to the mining nodes.  --> non-mining nodes don't matter.  Market cap is normal.

In case 2: miners will still continue to mine normal block awards.  Most non-mining full nodes, who updated their code, will reject the block chain miners produce.  If users send their transactions to mining or non-mining nodes, it doesn't matter, in the end these transactions will get caught by miners, who will mine it with normal rewards.  If users are pissed off because they didn't want this, they can simply sell their coins, and the market cap plunges.  --> non-mining nodes don' t matter, market cap plunges.

In case 3: same as case 1, except that this time, miners now mine high-reward blocks.  Same conclusion: non-mining nodes don't matter, market cap is normal.

In case 4: same as case 2, except that this time, miners now mine high-reward blocks.  Same conclusion: non-mining nodes don't matter, market cap plunges.

In other words: miners determine the protocol ; users determine the market cap.  Non-mining nodes don't matter.  When users are in disagreement with the miners, they leave, and the market cap plunges.  When the users are in agreement with the miners, both work together and keep up the transactions and the market cap.  Whatever think non-mining nodes doesn't matter.

Or if the Full Nodes detect that the coins have already been spent, it will deny its entry into the blockchain, (Preventing a double spend.)

The point is that no-one gives a shit about what block chain a non-mining node decides to accept or not, apart the owner of that node.  What matters is the chain on which a mining node works, and whether that node receives user transactions (directly) or not.

A block chain is a relationship between those writing the chain (the mining nodes) and those using it (the users).  The non-mining nodes can help in the transmission of the data, but in as much as they don't agree, it is as if they were simply switched off.

[dinofelis]

I would like to point to a few issues here:

Could PoW be done with an extremly low inflations, sure no dispute there.
However due to the economics involved it won't be, PoW mining energy & ASICS requirements costs are constantly growing for a successful PoW coin, because it is PoW it will become unaffordable.
PoS does not have the economic flaws of an ever increasing energy price or ASICS purchases, which is why it is the only method that will currently work.

In my opinion, there is absolutely no need for a GROWING mining cost for a PoW coin.   Ideally, the PoW cost is exactly equal to the *market value* of the coins being mined and the fees collected minus a small margin.  Let us concentrate on the coins being mined.  If the emission scheme is such that the number of coins emitted is not falling proportionally to the market appreciation of a single coin, then of course, the market value of the block reward increases, and HENCE the cost of PoW has to increase (as it has to be almost the reward value).  This has been the case with bitcoin: the market value of a block reward has increased over time, so the cost of mining had to increase too.  This is because the bitcoin emission scheme is TOO SLOW as compared to its market appreciation.  If the amount of coins mined dropped faster than the market appreciated a coin, then the mining cost would decrease, as it wouldn't be competitive to mine at higher cost than the block reward.

Suppose, for instance, that instead of a halving every 4 years, bitcoin had a halving every 4 months.  Then the amount of PoW waste right now would be way, way lower.  It isn't even said that there would have been ASICS.  It is simply so that the VALUE of a block reward is so high in bitcoin, that a lot of value has to be wasted in PoW.

On the other hand, if we would have a very slow emission scheme, where, say, the inflation rate would be 0.01% per year, then there too, there would be much less PoW cost: it would amount to the market cap divided by 10 000.  If it were bitcoin's market cap, it would result in a waste of about 1.5 million dollars a year at this point.   However, in as much as such a coin would appreciate a lot, even 0.01% of the market cap could mean a huge waste in the far future.

[alkan]

PoS as the consensus system is perfect,
the problem comes when you try to make a consensus system into a Interest Bearing Savings Account.

Indeed.

I tend to disagree. If the richest stakeholder will eventually own the whole currency (thanks to compound interests), then the consensus system will also become centralized.
In a standard PoS currency, wealth has a direct impact on consensus power. Decoupling stake and voting power seems possible though:
https://bitcointalk.org/index.php?topic=1702796.msg17071444#msg17071444
https://bitcointalk.org/index.php?topic=1719396.msg17216444#msg17216444

Another approach would be a negative interest scheme that requires people to mint blocks in order to not lose money over time. The number of blocks that a stakeholder needs to build per time period would depend on his stake.

[dinofelis]

PoS as the consensus system is perfect,
the problem comes when you try to make a consensus system into a Interest Bearing Savings Account.

Indeed.

I tend to disagree. If the richest stakeholder will eventually own the whole currency (thanks to compound interests), then the consensus system will also become centralized.

My "indeed" pointed to the problem, not to the "system is perfect" :-)

In fact, PoS would be a good system if there were NO REWARD (which is nothing else but an "interest bearing savings account").  But then it begs the question, because how do you do the coin emission then ?

Another approach would be a negative interest scheme that requires people to mint blocks in order to not lose money over time. The number of blocks that a stakeholder needs to build per time period would depend on his stake.

But again, how do you do coin emission ?  Where do the coins go that are "taxed" from the non-minting stake holders ?  Are they simply destroyed ?  (deflationary currency) ?  That makes the emission problem even worse: you start with no coins, and you have to destroy some ?

We are thinking here about "who takes the burden of securing the chain".  But that's just one problem.  The real monetary problem is "how do you do coin emission" ?  I have to say that I find PoW very well thought off in this respect: it solves both issues: it destroys seigniorage (so that coin emission is not "making a privileged rich" and hence renders that coin emission credible and acceptable) and at the same time it secures the chain. 
The downside is the centralisation that happens when too much gain is made through economies of scale, like bitcoin illustrates.  Another downside is that the PoW becomes insanely wasteful if the emission scheme doesn't correspond well to the coin appreciation, so that there are too important block rewards which must be burned (like with bitcoin).

I don't see how PoS can help with coin emission without becoming an "interest bearing savings account" which leads to terrible divergence due to the compound interest formula.  PoS CAN work to secure the chain if it is combined with some PoW, but that's not the hard part.  The hard part is coin emission.

[Flyskyhigh]

PoS works fine. MINT is a good example of this.

[klarki]

You may explain the advantage of simpler?

[kiklo]

Ask yourself what happens if:

1) all non-mining full nodes are in disagreement with a block
2) 70% of all mining nodes are in agreement with that block.

It depends,
if 70% of the mining ASICS agree the block award is 36 BTC per block, then the full nodes will block its entry into the blockchain.
Mining ASICS will fork off and Full Nodes will protect the BlockChain from corruption.  

And how would they do that ?  They will *observe* corruption, but what can they do about it ?  They cannot add any block.  They cannot make any forked chain (if they do, then they become mining nodes, which we supposed they weren't).

What does a non-mining, full node actually do, apart from verifying FOR HIMSELF whether the block chain that is distributed elsewhere actually corresponds to his own standards and wishes, or not ?  Once you see that the block chain that you downloaded from peers, doesn't correspond to what you thought it had to be, you can only refuse to use it, but that's about it.

Of course, when I talk about miners, I'm talking about the FULL NODES that mine (whether they are individual miners, or pools).  I'm not talking about JUST the mining hardware. I'm talking about the node that takes the decision on what block chain to build the next block.

Let us do a gedanken experiment.  Again, in what follows, the what I call a miner is the mining node, that decides on what block to work, on top of what chain.  THOSE are the (only) nodes that matter, as I will show.  A non-mining node (as I have one at home) doesn't have the slightest bit of "consensus power".

A fundamental modification of the protocol is proposed (like you say, the 36 BTC per block reward).  

We consider this:

EDIT: oops, I messed up, which made this incomprehensible.  Options corrected.

1) most miners are against it, but most non mining full nodes want to apply it, and users don't want want it.

2) most miners are against it, but most non mining full nodes want to apply it, and users want it.

3) most miners are in favor of it, but most non mining full nodes don't want it, and users want it.

4) most miners are in favor of it, but most non mining full nodes don't want it, and users don't want it.


What happens ?

In case 1, miners will still continue to mine normal block awards.  Most non-mining full nodes, who updated their code, will reject the block chain miners produce.  As users want it this way, they will short-circuit the non-mining nodes, and send their transactions directly to the mining nodes.  --> non-mining nodes don't matter.  Market cap is normal.

In case 2: miners will still continue to mine normal block awards.  Most non-mining full nodes, who updated their code, will reject the block chain miners produce.  If users send their transactions to mining or non-mining nodes, it doesn't matter, in the end these transactions will get caught by miners, who will mine it with normal rewards.  If users are pissed off because they didn't want this, they can simply sell their coins, and the market cap plunges.  --> non-mining nodes don' t matter, market cap plunges.

In case 3: same as case 1, except that this time, miners now mine high-reward blocks.  Same conclusion: non-mining nodes don't matter, market cap is normal.

In case 4: same as case 2, except that this time, miners now mine high-reward blocks.  Same conclusion: non-mining nodes don't matter, market cap plunges.

In other words: miners determine the protocol ; users determine the market cap.  Non-mining nodes don't matter.  When users are in disagreement with the miners, they leave, and the market cap plunges.  When the users are in agreement with the miners, both work together and keep up the transactions and the market cap.  Whatever think non-mining nodes doesn't matter.

Or if the Full Nodes detect that the coins have already been spent, it will deny its entry into the blockchain, (Preventing a double spend.)

The point is that no-one gives a shit about what block chain a non-mining node decides to accept or not, apart the owner of that node.  What matters is the chain on which a mining node works, and whether that node receives user transactions (directly) or not.

A block chain is a relationship between those writing the chain (the mining nodes) and those using it (the users).  The non-mining nodes can help in the transmission of the data, but in as much as they don't agree, it is as if they were simply switched off.

@Dinofelis,

Full Nodes perform a Validation of whether the Block meet the software code's criteria.
They Don't have to create any new blocks, to perform that function, what they do is Deny Blocks that fail the standards criteria.

The miner does the work, but the Full Nodes act as Quality Control, refusing blocks that Fail the Standards programed in the code,
you seem to have a blind spot on this.

So tell you what , Prove me wrong Generate a Block that has 15.5 BTC as a PoW reward instead of the 12.5 reward and get it accepted in the blockchain.
You can't do it because even if the miners agreed, the non mining Full Nodes will Block it because it fails the standard.
It keeps individuals from cheating the group.


 

[kiklo]

I don't see how PoS can help with coin emission without becoming an "interest bearing savings account" which leads to terrible divergence due to the compound interest formula.  PoS CAN work to secure the chain if it is combined with some PoW, but that's not the hard part.  The hard part is coin emission.

Ultra Low Inflation rate, of .0005% per year will stop a coin from being an interest bearing savings account,
even someone with 100 million coins will only be able to generate ~1.37 coin per day.

Adding a dead end like PoW to PoS is a waste of resources,
Every hybrid , the PoS becomes the dominate over PoW, because of the economics.