Author Topic: will you guys proofread my article about DPOS for me?  (Read 6198 times)

0 Members and 1 Guest are viewing this topic.

merockstar

  • Guest
No ability to vote against delegates.  Just for.

thats why i updated it. phrased it as people "remove their support"

Quote
If somebody were to attempt to do something harmful to the network, signing a bad block, failing to produce a block in a timely manner, or not including all the valid transactions that had been broadcast more than a minute before receiving the block, then the software automatically removes the support of people who haven't customized who they wish to vote for, and those choosing who they vote for will remove their support until the bad delegate is forced out of the position. The system can automatically fire any delegate if it detects proof that the delegate either signed multiple blocks at the same time, or declared a transaction to be included in the chain when it actually wasn't. In the event that something else of a major nature were to occur, individual clients also have the ability to go in and vote manually against any given delegate for a fee.

dont think I missed any of those references, did i?
« Last Edit: August 20, 2014, 03:59:56 am by merockstar »

Offline bytemaster

No ability to vote against delegates.  Just for.
For the latest updates checkout my blog: http://bytemaster.bitshares.org
Anything said on these forums does not constitute an intent to create a legal obligation or contract between myself and anyone else.   These are merely my opinions and I reserve the right to change them at any time.

merockstar

  • Guest
Updated:

Intro to Delegated Proof of Stake

Delegated Proof of Stake is a new method of securing a crypto-currency's network that attempts to solve some of the problems of Bitcoin's traditional Proof of Work system, and the Proof of Stake systems implemented by Peercoin and NXT, by implementing a layer of technological democracy to offset the negative effects of centralization.

At the heart of the creation of DPOS is the subject of centralization. With Proof of Work, centralization just kind of crept it's way in over time along with the advent of computers designed specifically for mining called ASICs. The production of specialized computers for this purpose is limited to a segment of the population with the resources to design and produce computer parts. The tremendous amount of hash power that could be produced by these machines dwarfed the GPU and FPGA mining rigs that had been previously securing the network, and the end result was that a small number of miners, who already had an economic advantage, ended up securing most of the network. If they were to collaborate into mining pools, this gives these mining pools the ability to combine resources and gain over fifty one percent of the network hash rate. Having that much of the hashrate gives these mining pools the ability to falsify the blockchain if they wanted to, meaning the greater community has to trust that they won't do that.

Centralization seems to be an inevitability in cryptocurrency. Even with Proof of Stake the cost of accumulating enough network resources to falsify the blockchain long enough to create a double spend is much higher, but not out of the realm of possibility. Instead of making blocks more computationally difficult to produce, proof of stake weighs the amount of coin being held by each individual in the network, and chooses who will sign blocks from a pool of people who have the most to lose from the network being compromised.

Delegated proof of stake doesn't deny the fact that centralization is an inevitability, but instead mitigates the potential negative impacts of this by spreading out the job of signing blocks to an optimum number of delegates who are voted on by everybody using the network with every transaction that gets made. This makes this inevitable bit of centralization more democratic in nature with a decentralized voting process. Instead of trying to reduce the need for trust all together, DPOS has safeguards in place that ensure that those trusted with signing blocks on behalf of the network are doing their job correctly. Additionally, each block that gets signed must have verification that the block before it got signed by a trusted node. This eliminates the need in other types of blockchains to have to wait until a certain number of untrusted nodes have verified it for confirmation.

This reduced need for confirmation produces a dramatic increase in speed of transaction times, and this new system of intentionally placing trust with the most trustworthy of potential block signers, as decided by the network, has the added benefit that no artificial encumbrance need be imposed to slow down the block signing process. This allows for many more transactions to be included in a block than in traditional proof of work and proof of stake systems. This brings cryptocurrency to a level where it may be able to compete with the centralized clearinghouses like Visa and Mastercard who administer the currently most popular forms of electronic payment systems in the world.

Under DPOS each of the trusted delegates running nodes receive a certain amount of pay coming from the transaction fees in the blocks that they sign. Depending on the needs of the DAC, and the needs of the delegate, the amount received can be flexible. At the delegate's discretion this pay can be compensation, or can be used to further the DAC by means of marketing, development, or whatever else might be required. How much of these funds are drawn, and what they are spent on can be used as campaign points for getting/staying elected.

The compensatory part of this pay has the effect of making the job sought after, and incentivises elected delegates to keep the job by showing the rest of the network that they are doing it properly, using the standard software. To be able to disrupt the network in any way you would first have to secure the position of delegate by being elected by the network, then, once you're in, the potential for profit from doing this job correctly outweighs the potential to profit from attempting to disrupt the network in anyway.

If somebody were to attempt to do something harmful to the network, signing a bad block, failing to produce a block in a timely manner, or not including all the valid transactions that had been broadcast more than a minute before receiving the block, then the software automatically removes the support of people who haven't customized who they wish to vote for, and those choosing who they vote for will remove their support until the bad delegate is forced out of the position. The system can automatically fire any delegate if it detects proof that the delegate either signed multiple blocks at the same time, or declared a transaction to be included in the chain when it actually wasn't. In the event that something else of a major nature were to occur, individual clients also have the ability to go in and vote manually against any given delegate for a fee.

In a delegated proof of stake system centralization still occurs, but it is controlled. Unlike with other methods of securing a cryptocurrency network, every client in a DPOS system gets the ability to decide who is trusted with this centralization instead of this trust naturally finding it's way to those with the most resources. This allows the network to reap some of the major advantages of centralization, while still maintaining a decentralized nature to a certain calculated extent, enforced by a fair election process where anyone could potentially become a delegated representative of the majority of users.

Offline onceuponatime


merockstar

  • Guest
Updated it above to reflect onceuponatime's suggestions.

Offline onceuponatime

[quote author=merockstar link=topic=5308.msg70905#msg70905

Delegated proof of stake doesn't deny the fact that centralization is an inevitability, but instead mitigates the potential negative impacts of this by spreading out the job of signing blocks to 101 delegates who are voted on by everybody using the network with every transaction that gets made. This makes this inevitable bit of centralization more democratic in nature with a decentralized voting process. Instead of trying to reduce the need for trust all together, DPOS has safeguards in place the (that)ensure that those trusted with signing blocks on behalf of the network are doing their job correctly. Additionally, each block that gets signed must have verification that the block before it got signed by a trusted node. Eliminating the need in other types of blockchains to have to wait until a certain number of untrusted nodes have verified it for confirmation.

This reduced need for confirmation produces an increase in speed of transaction times, and this new system of intentionally placing trust with the most trustworthy of potential block signers, as decided by the network, has the added benefit that no artificial encumbrance need be imposed to slow down the block signing process. Allowing(This allows) for many more transactions to be included in a block than in traditional proof of work and proof of stake systems. This brings cryptocurrency to a level where it can compete with the centralized clearinghouses like Visa and Mastercard who administer the currently most popular forms of electronic payment systems in the world.

Under DPOS each of the trusted delegates running nodes receives (receive) a certain amount of pay. Depending on the needs of the DAC, and the needs of the delegate, the amount received can be flexible. It can come from a certain percentage of transaction fees from each block that the delegate signs, or if more is required could also come from block rewards. At the delegate's discretion this pay can be compensation, or can be used to further the DAC by means of marketing, development, or whatever else might be required. How much of these funds are drawn, and what they are spent on can be used as campaign points.

The compensatory part of this pay has the effect of making the job sought after, and incentivises them to keep the job by showing the rest of the network they they are doing it properly, using the standard software. To be able to disrupt the network in any way you would first have to secure the position of delegate by being elected by the network, then once you're in, the potential for profit from doing this job correctly outweighs the potential to profit from attempting to disrupt the network in anyway (any way).

[/quote]

This is a first class article.  :)

Offline bytemaster

For the latest updates checkout my blog: http://bytemaster.bitshares.org
Anything said on these forums does not constitute an intent to create a legal obligation or contract between myself and anyone else.   These are merely my opinions and I reserve the right to change them at any time.

Offline xeroc

  • Board Moderator
  • Hero Member
  • *****
  • Posts: 12922
  • ChainSquad GmbH
    • View Profile
    • ChainSquad GmbH
  • BitShares: xeroc
  • GitHub: xeroc

merockstar

  • Guest
second (possibly final) draft

I'm pretty happy with this right now. So I'm going to call it done unless somebody has any more input.




I've added a few commas, and a couple of grammatical tweaks and a few words that may add to clarity in your extremely well done article. See if you want to use any of the changes:

Intro to Delegated Proof of Stake

Delegated Proof of Stake is a new method of securing a crypto-currency's network that attempts to solve some of the problems of Bitcoin's traditional Proof of Work system, and the Proof of Stake systems implemented by Peercoin and NXT, by implementing a layer of technological democracy to offset the negative effects of centralization.

At the heart of the creation of DPOS is the subject of centralization. With Proof of Work, centralization just kind of crept it's way in over time along with the advent of computers designed specifically for mining called ASICs. The production of specialized computers for this purpose is limited to a segment of the population with the resources to design and produce computer parts. The tremendous amount of hash power that could be produced by these machines dwarfed the GPU and FPGA mining rigs that had been previously securing the network, and the end result was that a small number of miners, who already had an economic advantage, ended up securing most of the network. If they were to collaborate into mining pools, this gives these mining pools the ability to combine resources and gain over fifty one percent of the network hash rate. Having that much of the hashrate gives these mining pools the ability to falsify a blockchain if they wanted to, meaning the greater community has to trust that they won't do that.

Centralization seems to be an inevitability in cryptocurrency. Even with Proof of Stake the cost of accumulating enough network resources to falsify the blockchain long enough to create a double spend is much higher, but not out of the realm of possibility. Instead of making blocks more computationally difficult to produce, proof of stake weighs the amount of coin being held by each individual in the network, and chooses who will sign blocks by who has the most to lose from the network being compromised.

Delegated proof of stake doesn't deny the fact that centralization is an inevitability, but instead mitigates the potential negative impacts of this by spreading out the job of signing blocks to 101 delegates who are voted on by everybody using the network with every transaction that gets made. This makes this inevitable bit of centralization more democratic in nature with a decentralized voting process. Instead of trying to reduce the need for trust all together, DPOS has safeguards in place that ensure that those trusted with signing blocks on behalf of the network are doing their job correctly. Additionally, each block that gets signed must have verification that the block before it got signed by a trusted node. This eliminates the need in other types of blockchains to have to wait until a certain number of untrusted nodes have verified it for confirmation.

This reduced need for confirmation produces an increase in speed of transaction times, and this new system of intentionally placing trust with the most trustworthy of potential block signers, as decided by the network, has the added benefit that no artificial encumbrance need be imposed to slow down the block signing process. This allows for many more transactions to be included in a block than in traditional proof of work and proof of stake systems. This brings cryptocurrency to a level where it can compete with the centralized clearinghouses like Visa and Mastercard who administer the currently most popular forms of electronic payment systems in the world.

Under DPOS each of the trusted delegates running nodes receive a certain amount of pay. Depending on the needs of the DAC, and the needs of the delegate, the amount received can be flexible. It can come from a certain percentage of transaction fees from each block that the delegate signs, or, if more funds are required, could also come from block rewards. At the delegate's discretion this pay can be compensation, or can be used to further the DAC by means of marketing, development, or whatever else might be required. How much of these funds are drawn, and what they are spent on, can be used as campaign points for getting/staying elected.

The compensatory part of this pay has the effect of making the job sought after, and incentivises elected delegates to keep the job by showing the rest of the network that they are doing it properly, using the standard software. To be able to disrupt the network in any way you would first have to secure the position of delegate by being elected by the network, then, once you're in, the potential for profit from doing this job correctly outweighs the potential to profit from attempting to disrupt the network in anyway.

If somebody were to attempt to do something harmful to the network, signing a bad block, failing to produce a block in a timely manner, or not including all the valid transactions that had been broadcast more than a minute before receiving the block, then the software automatically causes all the running clients to start voting against that delegate until they are forced out of the position. The system can automatically fire any delegate if somebody were to produce proof that the delegate either signed multiple blocks at the same time, or declared a transaction to be included in the chain when it actually wasn't. In the event that something else of a major nature were to occur, individual clients also have the ability to go in and vote manually against any given delegate for a fee.

In a delegated proof of stake system centralization still occurs, but it is controlled. Unlike with other methods of securing a cryptocurrency network, every client in a DPOS system gets the ability to decide who is trusted with this centralization instead of this trust naturally finding it's way to those with the most resources. This allows the network to reap some of the major advantages of centralization, while still maintaining a decentralized nature to a certain calculated extent, enforced by a fair election process where anyone could potentially become a delegated representative of the majority of users.


EDIT: updated to reflect input from onceuponatime
« Last Edit: June 28, 2014, 02:54:09 am by merockstar »

merockstar

  • Guest

Delegates can only be fired when someone produces a transaction with crypto-graphic proof that a delegate has signed a false statement (multiple blocks for the same time, or declared a transaction to be included in the chain when it wasn't).   


Be automatically fired or by approval removed manually.

From the whitepaper:

Quote
When bad behavior is detected by any delegate all clients will automatically vote against that delegate the next time their user makes a transaction until that delegate is no longer in the top 200.

sorry I overlooked that for so long heyD

Offline Simeon II

  • Sr. Member
  • ****
  • Posts: 355
    • View Profile
but... the house of representin' (idiocracy reference) is weighted by the population of each state. maybe I could explain that it is democratic in a house way, but not in a senate way.


If you take advice from newbies:

This will be going too much country specific. Keep in mind that:

1. Every country is deferent; 2. US is deferent election-wise from most everybody.

2c.


Offline luckybit

  • Hero Member
  • *****
  • Posts: 2921
    • View Profile
  • BitShares: Luckybit
^^ Actually we are pretty vehemently anti-democracy for these sorts of systems. Democracy means one vote per person, which is not how DACs work.

Toast is right that we are rather against the "one-vote, one person" perspective, but I also know that people like toast and I think differently than the masses in that we demand accurate descriptions and hate loose words.    That said, though democracy isn't a true ideal to be strived for, it is something that many people have associated with fairness. 

So I suggest we simply redefine democracy and democratic to suite our purposes and gain the good will that 99% of the population associates with the term.   Lets embrace the positive propaganda rather than fight it and then include in fine print in a blog article the nuance interpretation that keeps it consistant with our ideals.

In my opinion that would open us up to attack. Fine print is a bad way to mitigate the fact that we are using a word unconventionally/falsely to gain emotional appeal from the unaware - and we would surely be caught out and attacked.

I don't think we are using it falsely.   Most people would say that companies operate on a the principle of democracy among shareholders.   Democracy is nothing more than majority rules and it is still democratic when you weight votes by shares.   

I think the issue is whether or not a democratic process is considered "good" and I have argued that it is inherently "bad" because the minority has their property rights subjected to the majority.    IE: a democracy is a way of fighting without bloodshed.  It is not a valid means to reach an honest answer about what is right or wrong.

The premise of democracy is that the majority is "good" and thus it should prevent bad actors from gaining control.   I think this may be generally true in systems with an enforceable constitution such as blockchains, but is clearly flawed in governments.

You are preaching to the choir  :)   My only concern is being open to attack by manipulators of public opinion. (Not everyone has your ethical foundation - and if something can be twisted it will be).

With framing and marketing its better to always be on the offensive rather than defensive. Look at elections and tell me what happens with the candidate that is more concerned about being attacked? Do those sorts of candidates ever win an election?
https://metaexchange.info | Bitcoin<->Altcoin exchange | Instant | Safe | Low spreads

Offline luckybit

  • Hero Member
  • *****
  • Posts: 2921
    • View Profile
  • BitShares: Luckybit
^^ Actually we are pretty vehemently anti-democracy for these sorts of systems. Democracy means one vote per person, which is not how DACs work.

Toast is right that we are rather against the "one-vote, one person" perspective, but I also know that people like toast and I think differently than the masses in that we demand accurate descriptions and hate loose words.    That said, though democracy isn't a true ideal to be strived for, it is something that many people have associated with fairness. 

So I suggest we simply redefine democracy and democratic to suite our purposes and gain the good will that 99% of the population associates with the term.   Lets embrace the positive propaganda rather than fight it and then include in fine print in a blog article the nuance interpretation that keeps it consistant with our ideals.

In my opinion that would open us up to attack. Fine print is a bad way to mitigate the fact that we are using a word unconventionally/falsely to gain emotional appeal from the unaware - and we would surely be caught out and attacked.

Satoshi said One-CPU-one-Vote.

So all along these technologies were intended to at least resemble a democracy. Whether or not it is actually a democracy is something more of mathematics but as long as anyone can become a delegate it's a lot more fair than what we have now.

https://metaexchange.info | Bitcoin<->Altcoin exchange | Instant | Safe | Low spreads

Offline onceuponatime

^^ Actually we are pretty vehemently anti-democracy for these sorts of systems. Democracy means one vote per person, which is not how DACs work.

Toast is right that we are rather against the "one-vote, one person" perspective, but I also know that people like toast and I think differently than the masses in that we demand accurate descriptions and hate loose words.    That said, though democracy isn't a true ideal to be strived for, it is something that many people have associated with fairness. 

So I suggest we simply redefine democracy and democratic to suite our purposes and gain the good will that 99% of the population associates with the term.   Lets embrace the positive propaganda rather than fight it and then include in fine print in a blog article the nuance interpretation that keeps it consistant with our ideals.

In my opinion that would open us up to attack. Fine print is a bad way to mitigate the fact that we are using a word unconventionally/falsely to gain emotional appeal from the unaware - and we would surely be caught out and attacked.

I don't think we are using it falsely.   Most people would say that companies operate on a the principle of democracy among shareholders.   Democracy is nothing more than majority rules and it is still democratic when you weight votes by shares.   

I think the issue is whether or not a democratic process is considered "good" and I have argued that it is inherently "bad" because the minority has their property rights subjected to the majority.    IE: a democracy is a way of fighting without bloodshed.  It is not a valid means to reach an honest answer about what is right or wrong.

The premise of democracy is that the majority is "good" and thus it should prevent bad actors from gaining control.   I think this may be generally true in systems with an enforceable constitution such as blockchains, but is clearly flawed in governments.

You are preaching to the choir  :)   My only concern is being open to attack by manipulators of public opinion. (Not everyone has your ethical foundation - and if something can be twisted it will be).

merockstar

  • Guest

Too late - I published it before I saw this.
Let me know and I'll replace it as soon as its ready.

http://bitshares.org/intro-to-delegated-proof-of-stake/

suuuuuure. lol.

alrighty. :)

note to self: reiterate that even though a vote happens every transaction, the procedure is pretty much automatic unless it need not be
« Last Edit: June 26, 2014, 05:04:07 am by merockstar »