Topic: DECENTRALIZED crypto currency (including Bitcoin) is a delusion (any solutions?) - page 23. (Read 91144 times)
July 13, 2016, 03:41:00 PM
July 07, 2016, 05:28:51 PM
I don't have time to read all of your posts.
And I also don't have time to spoon feed you. You'll know at the appropriate time why you were incorrect. Until then, enjoy your beliefs.
July 07, 2016, 05:26:07 PM
Electricity and equipment are real costs that need to be paid for. Unfortunately, those costs are often too high for transaction fees to be enough.
Given the obligation of the validators to do the unprofitable proof-of-work is vacated, the remaining electricity costs approach epsilon.
CoinHoarder has written nonsense, but I will not waste my time rebuking him now. I will rebuke him in a comprehensive manner at the appropriate time in the future, although the main issue with DPoS was recently explained to him. I don't intend this to be a personal insult. I am just speaking factually. Nonsense is nonsense. Note I am not speaking to his legal point, because he knows damn well that I had a major epiphany in my legal interpretation since the time of the posts that he is quoting.
I am sure everyone can decide for themselves what is nonsense and what is not. Unlike you, I am able to convey my thoughts in an easy to understand manner without linking to walls of text and hundred page threads. I am sure most people can easily understand my objections to your arguments.
I still maintain that you are pumping your vaporware. I just retreated from that thread because you kindly asked me to.
Also, just because you posted something does not mean I am aware that you did so. I don't have time to read all of your posts. You are still linking to the post that lists your qualms with PoS as if you still maintain those qualms, so excuse me if you have changed your mind.
July 07, 2016, 04:10:44 PM
Electricity and equipment are real costs that need to be paid for. Unfortunately, those costs are often too high for transaction fees to be enough.
Given the obligation of the validators to do the unprofitable proof-of-work is vacated, the remaining electricity costs approach epsilon.
CoinHoarder has written nonsense, but I will not waste my time rebuking him now. I will rebuke him in a comprehensive manner at the appropriate time in the future, although the main issue with DPoS was recently explained to him. I don't intend this to be a personal insult. I am just speaking factually. Nonsense is nonsense. Note I am not speaking to his legal point, because he knows damn well that I had a major epiphany in my legal interpretation since the time of the posts that he is quoting.
July 07, 2016, 03:23:48 PM
PoS usually pays dividends to stake holders (and even relays a percentage to the developers thus must register as a Money Transmitter with FinCEN) thus arguably creating investment securities under the Howey test and thus must be registered with the SEC or face possible jail time. I argue this impacts the resilience.
This might apply to most forms of PoS, but dPoS destroys transaction fees. I don't think that you can count that as having paid dividends to shareholders, as there are no transactions to each shareholder doing so. Yes, destroying transaction fees increases shareholder's percentage of the pie, but I think it is a bit of a stretch to claim it is the same as paying dividends to shareholders.
The other flaw of PoS, and especially DPOS and Dash masternodes (as pointed out by smooth et al) is you are paying yourselves via the shares from an enterprise that issued unregistered investment securities and which also requires each stakeholder to register as a money transmitter with FinCIN. I can't fathom how you convinced yourself that you are not going to jail in the future or end having to lick the boots of the SEC as Erik Voorhees did to wiggle out of jail time.
1. Again, dPoS is not paying dividends.
2. This point is conjecture from "armchair forum lawyers" (read: not lawyers). Obviously, most cryptocurrencies operate in a grey area, but to claim it is black and white like that is stretching the truth.
July 07, 2016, 03:02:00 PM
centralizes control according to stake, which is a finite resource
....
there is no way to distribute new coins (must distribute proportional to stake in order to be fair thus effectively no change in coin distribution)
....
there is no way to distribute new coins (must distribute proportional to stake in order to be fair thus effectively no change in coin distribution)
You still are under the misconception that stake is a finite resource within PoS cryptocurrencies.
By marrying PoS and PoW it is possible:
A. Far advance warning of an cryptocurrencies' distribution/fundraising, with a clear objective and roadmap for the project and a set timeframe for the distribution/fundraising.
B. A working closed source beta, to be open-sourced after the original distribution period
C. PoW distribution (50% of the genesis stake)
Ca. Use several different mining algorithms, with each having an their own separate difficulty and the same chance to find the next block (implemented in several coins.. MyriadCoin was the first)
Cb. Use at least one algorithm per different types of mining hardware (one for Scrypt ASICs, one for SHA256 ASICs, one for CPUs, one for ATI GPUs, one for NVIDIA GPUs, etc..)
Cc. The point of this is to allow people with all types of minin hardware to participate so it is inclusive of everyone.
D. A regular IPO distribution (50% of the genesis stake)
Da. This is to allow people to participate in the initial distribution that do not want to hassle with mining equipment.
E. Launch said closed source beta into the wild after the distribution/fund-raising has occurred and switch to PoS
F. To mitigate claims of having an unfair PoS distribution to early insiders...
Fa. Have an annual block-limited PoW period that issues more coins to PoW miners.
Fb. The amount of tokens mined in the annual PoW periods should be reduced each year until eventually there is only a small amount of coins issued annually in perpetuity (aka. a tail emission).
Fc. This is intended to mimic the block reduction scheme and distribution of most PoW currencies, but with an added tail emission.
B. A working closed source beta, to be open-sourced after the original distribution period
C. PoW distribution (50% of the genesis stake)
Ca. Use several different mining algorithms, with each having an their own separate difficulty and the same chance to find the next block (implemented in several coins.. MyriadCoin was the first)
Cb. Use at least one algorithm per different types of mining hardware (one for Scrypt ASICs, one for SHA256 ASICs, one for CPUs, one for ATI GPUs, one for NVIDIA GPUs, etc..)
Cc. The point of this is to allow people with all types of minin hardware to participate so it is inclusive of everyone.
D. A regular IPO distribution (50% of the genesis stake)
Da. This is to allow people to participate in the initial distribution that do not want to hassle with mining equipment.
E. Launch said closed source beta into the wild after the distribution/fund-raising has occurred and switch to PoS
F. To mitigate claims of having an unfair PoS distribution to early insiders...
Fa. Have an annual block-limited PoW period that issues more coins to PoW miners.
Fb. The amount of tokens mined in the annual PoW periods should be reduced each year until eventually there is only a small amount of coins issued annually in perpetuity (aka. a tail emission).
Fc. This is intended to mimic the block reduction scheme and distribution of most PoW currencies, but with an added tail emission.
I am less certain about this, but I think that if the parameters/economics of the above are designed sufficiently then it can even thwart this gripe as well:
attacking the coin is a one-time cost of stake that sustains forever, whereas for Proof-of-Work the attacker must continue to expend resources on mining to maintain an attack
Since the attacker would have to compete during the annual PoW phases to retain control of the PoS cryptocurrency, it can no longer be considered a "one-time cost".
July 07, 2016, 02:45:25 PM
July 07, 2016, 01:58:25 PM
I don't speak Esperanto.
Your idea of a shared economic incentive is interesting, but it has to be enough to outweigh the costs of mining or in other words, hosting the register and handling transactions. Electricity and equipment are real costs that need to be paid for. Unfortunately, those costs are often too high for transaction fees to be enough. (Especially with an infant currency)
Of course, you'll have "believers" who will host it at a loss, but those decay unless use eventually provides the financial incentive to make it worthwhile. Depending on how bandwidth/power/instruction hungry your process is will determine what that break even point would be for someone.
Even with slavery, someone is still paying for the gold miners to mine the gold.
Your idea of a shared economic incentive is interesting, but it has to be enough to outweigh the costs of mining or in other words, hosting the register and handling transactions. Electricity and equipment are real costs that need to be paid for. Unfortunately, those costs are often too high for transaction fees to be enough. (Especially with an infant currency)
Of course, you'll have "believers" who will host it at a loss, but those decay unless use eventually provides the financial incentive to make it worthwhile. Depending on how bandwidth/power/instruction hungry your process is will determine what that break even point would be for someone.
Even with slavery, someone is still paying for the gold miners to mine the gold.
July 07, 2016, 08:47:23 AM
Yet without profit, you will hinder adoption.
If miners-for-profit are the main adopters of a token, then you don't really have a mass adopted currency.
The main problem to solve with unprofitable mining is to prevent the centralization of validation of the block chain, to do so without introducing other game theories for profit which destroy the Nash equilibrium, and to achieve this with a design that can scale to any level of transactions per second.
In profitable mining, every miner has an economic incentive to validate the block chain, lest the miner lose the block reward (coinbase+txn fees) because other miners will not mine on an invalid block.
In unprofitable mining, there is an analogous economic incentive to validate the block chain axammonChaxaupp disttnaxaupp envaxaupp axaronobboktonShaxaupp dinyaxaupp oldaxaupp olgfaxauictuchlioFaxaupp . That is the closest hint I have ever provided as to the correct design for unprofitable mining. One of the keys to achieving that is a shared economic incentive that binds all nodes to a Nash equilibrium.
July 06, 2016, 11:01:35 AM
So long as mining is profitable, it will centralize to the person/organization/country who has the means to profit the most from it. Yet without profit, you will hinder adoption.
It is a challenge to prevent centralization and be able to provide consensus rapidly enough for transactions to make sense for the consumer.
It's a difficult model. You can (likely) solve the problem in code, but the other half of it is the non-code adoption of the currency by the general public in order to make it decentralized and effectively hosted so that the ledger/currency can exist and function. So you have to incentivize. And thus, you create centralization.
I don't know how you solve this. Smart people are constantly thinking about it. TPBB has an idea that may or may not solve it. Still, the algo solution may not solve the issue with adoption. You have to have both if you want your currency to thrive.
It is a challenge to prevent centralization and be able to provide consensus rapidly enough for transactions to make sense for the consumer.
It's a difficult model. You can (likely) solve the problem in code, but the other half of it is the non-code adoption of the currency by the general public in order to make it decentralized and effectively hosted so that the ledger/currency can exist and function. So you have to incentivize. And thus, you create centralization.
I don't know how you solve this. Smart people are constantly thinking about it. TPBB has an idea that may or may not solve it. Still, the algo solution may not solve the issue with adoption. You have to have both if you want your currency to thrive.
July 03, 2016, 04:30:49 PM
Solution is already built in:
https://bitcointalksearch.org/topic/simple-solution-to-the-problem-of-centralisation-in-btc-already-built-in-1536282
The solution is: dump to double digits because it's centralised and wait for everyone to liquidate their industrial mining. Market cleaned out, all good.
We need to flush the miningsector! It will occure naturally.
https://bitcointalksearch.org/topic/simple-solution-to-the-problem-of-centralisation-in-btc-already-built-in-1536282
The solution is: dump to double digits because it's centralised and wait for everyone to liquidate their industrial mining. Market cleaned out, all good.
We need to flush the miningsector! It will occure naturally.
Lol. Flush your nonsense.
July 03, 2016, 12:32:45 PM
Solution is already built in:
https://bitcointalksearch.org/topic/simple-solution-to-the-problem-of-centralisation-in-btc-already-built-in-1536282
The solution is: dump to double digits because it's centralised and wait for everyone to liquidate their industrial mining. Market cleaned out, all good.
We need to flush the miningsector! It will occure naturally.
https://bitcointalksearch.org/topic/simple-solution-to-the-problem-of-centralisation-in-btc-already-built-in-1536282
The solution is: dump to double digits because it's centralised and wait for everyone to liquidate their industrial mining. Market cleaned out, all good.
We need to flush the miningsector! It will occure naturally.
July 03, 2016, 11:03:09 AM
We were doing good before asics were made. We could do alot by looking at algos that allow more of the public to participate in mining on common devices, instead of requiring dedicated equipment that makes mining more a business endeavor than a casual pursuit at this point.
July 03, 2016, 08:58:41 AM
I have some ideas for this. I doubt it is possible to have a single big decentealised cryptocurrency but with many different cryptocurrencies decentralisation is possible(if one get killed the others will survive).
I do not view "unprofitable proof of work" as a viable solution, if it is unprofitable to mine the coin it will just ve easier for a government to kill it. You can decrease the performance gap between CPU:s, GPU:s and ASIC but you cannot eliminate it.
My idea is to use the masternode system where the block reward decreases exponentially with the number of masternodes, with that system there will be a incentive for a person/group not to own too many masternodes.
Another idea i have is to having fixed nodes controlling the network(999 nodes with one votes each), than the currency will be controlled by elite, you may not like it but maybe it is the only viable solution.
I do not view "unprofitable proof of work" as a viable solution, if it is unprofitable to mine the coin it will just ve easier for a government to kill it. You can decrease the performance gap between CPU:s, GPU:s and ASIC but you cannot eliminate it.
My idea is to use the masternode system where the block reward decreases exponentially with the number of masternodes, with that system there will be a incentive for a person/group not to own too many masternodes.
Another idea i have is to having fixed nodes controlling the network(999 nodes with one votes each), than the currency will be controlled by elite, you may not like it but maybe it is the only viable solution.
June 16, 2016, 09:30:31 PM
wbb has the answer 
coming soon
coming soon
June 16, 2016, 12:34:27 PM
Steemit's PoW has a novel idea to incorporate signing into the PoW for the public key that will receive the coinbase block reward, so as to force the miner to have real-time access to the private key for each nonce attempted.
Daniel Larimer/Bitshares et. al claim some benefits which are explained at the following references:
https://steem.io/documentation/consensus/#mining-algorithm
https://steem.io/SteemWhitePaper.pdf#page=23
This is perhaps the first technical idea I've seen from the Dan Larimer group that I must frankly state is very good and I will likely adopt it for my own project. However, I would structure it differently to avoid the security caveats they noted. I would have the miner sign the hash(H, nonce), the use that as input to the PoW, e.g. PoW = hash(H,nonce,SIGN(hash(H, nonce))).
They think one of the benefits is that it will encourage the development of an ASIC which can perform faster elliptic curve signing and validation, and I presume this is so that they hope to be able to validate more transactions per second with more efficient hardware. But I don't claim that is a realistic expectation for a benefit, because well I expect to popularize unprofitable mining and thus the end of ASICs for mining forever.
The benefits I think are realistic are no botnets and no pools. Excellent. I had already eliminated farming out mining with another method (farming out can't be resistant to DDoS), and this idea from Dan's group completes attributes I'd like mining to have.
So for the first time, I need to tip my hat to Dan and kudos.
Daniel Larimer/Bitshares et. al claim some benefits which are explained at the following references:
https://steem.io/documentation/consensus/#mining-algorithm
https://steem.io/SteemWhitePaper.pdf#page=23
This is perhaps the first technical idea I've seen from the Dan Larimer group that I must frankly state is very good and I will likely adopt it for my own project. However, I would structure it differently to avoid the security caveats they noted. I would have the miner sign the hash(H, nonce), the use that as input to the PoW, e.g. PoW = hash(H,nonce,SIGN(hash(H, nonce))).
They think one of the benefits is that it will encourage the development of an ASIC which can perform faster elliptic curve signing and validation, and I presume this is so that they hope to be able to validate more transactions per second with more efficient hardware. But I don't claim that is a realistic expectation for a benefit, because well I expect to popularize unprofitable mining and thus the end of ASICs for mining forever.
The benefits I think are realistic are no botnets and no pools. Excellent. I had already eliminated farming out mining with another method (farming out can't be resistant to DDoS), and this idea from Dan's group completes attributes I'd like mining to have.
So for the first time, I need to tip my hat to Dan and kudos.
I remembered I had read something about non-outsourceable PoW in the past (perhaps also on Vitalik's blog) but I remember it has being more complex and unworkable so someone helped me find these old resources which I also remember having seen before and there is proof because I posted there (first one well before any possible sighting on Vitalik's Ethereum blog):
https://bitcointalk.org/index.php?topic=309073.20
https://bitslog.wordpress.com/2014/06/19/theoretical-and-practical-nonoutsourceable-puzzles/
The first one from Andrew Miller makes the valid point that if you kill all pools, then you force consolidation (centralization) of mining hashrate into to mining farms in order to control variance costs. But luckily that is an inapplicable issue in my plans for unprofitable mining.
The second links to a proposal I don't remember seeing before which is more similar to the one proposed for STEEM, which Andrew Miller also commented on:
http://hackingdistributed.com/2014/06/18/how-to-disincentivize-large-bitcoin-mining-pools/
And under the §Related Work, it links to:
http://hackingdistributed.com/2014/06/13/time-for-a-hard-bitcoin-fork/#comment-1435809647
That links to:
https://bitcointalksearch.org/topic/preventing-pool-mining-652443
Which appears to be exactly the same as the improvement I have suggested in the quoted post above.
June 16, 2016, 06:02:11 AM
And I found recently CIYAM speaking about a new 'proof-of-storage' here where I wonder if you could force now the miners also to have the entire blockchain stored locally for real:
https://bitcointalksearch.org/topic/m.15091671
https://bitcointalksearch.org/topic/m.15091671
In the post you linked, CIYAM apparently hadn't yet realized the insoluble Sybil attack problem for decentralized file storage, which I had originally pointed out in 2013, reiterated months ago, and reiterated again this week:
No decentralized storage can solve the Sybil attack on storage where many nodes are sharing the same storage, but collecting payments or otherwise deceiving the claimed redundancy as if they have multiple copies of the storage.
It is a fundamental weakness that can not be fixed. Decentralized storage can not work. Period. End of story.
Latency is not reliable enough to be used as an indicator.
It is a fundamental weakness that can not be fixed. Decentralized storage can not work. Period. End of story.
Latency is not reliable enough to be used as an indicator.
I hope I don't have to repeat that again.
AnonyMint's very first attempt at a better consensus algorithm was proof-of-diskspace in 2013.
Was not aware of that - tnx.
June 16, 2016, 05:59:17 AM
And I found recently CIYAM speaking about a new 'proof-of-storage' here where I wonder if you could force now the miners also to have the entire blockchain stored locally for real:
https://bitcointalksearch.org/topic/m.15091671
https://bitcointalksearch.org/topic/m.15091671
In the post you linked, CIYAM apparently hadn't yet realized the insoluble Sybil attack problem for decentralized file storage, which I had originally pointed out in 2013, reiterated months ago, and reiterated again this week:
No decentralized storage can solve the Sybil attack on storage where many nodes are sharing the same storage, but collecting payments or otherwise deceiving the claimed redundancy as if they have multiple copies of the storage.
It is a fundamental weakness that can not be fixed. Decentralized storage can not work. Period. End of story.
Latency is not reliable enough to be used as an indicator.
It is a fundamental weakness that can not be fixed. Decentralized storage can not work. Period. End of story.
Latency is not reliable enough to be used as an indicator.
I hope I don't have to repeat that again.
AnonyMint's very first attempt at a better consensus algorithm was proof-of-diskspace in 2013.
June 16, 2016, 05:03:45 AM
Steemit's PoW has a novel idea to incorporate signing into the PoW for the public key that will receive the coinbase block reward, so as to force the miner to have real-time access to the private key for each nonce attempted.
Daniel Larimer/Bitshares et. al claim some benefits which are explained at the following references:
https://steem.io/documentation/consensus/#mining-algorithm
https://steem.io/SteemWhitePaper.pdf#page=23
This is perhaps the first technical idea I've seen from the Dan Larimer group that I must frankly state is very good and I will likely adopt it for my own project. However, I would structure it differently to avoid the security caveats they noted. I would have the miner sign the hash(H, nonce), the use that as input to the PoW, e.g. PoW = hash(H,nonce,SIGN(hash(H, nonce))).
They think one of the benefits is that it will encourage the development of an ASIC which can perform faster elliptic curve signing and validation, and I presume this is so that they hope to be able to validate more transactions per second with more efficient hardware. But I don't claim that is a realistic expectation for a benefit, because well I expect to popularize unprofitable mining and thus the end of ASICs for mining forever.
The benefits I think are realistic are no botnets and no pools. Excellent. I had already eliminated farming out mining with another method (farming out can't be resistant to DDoS), and this idea from Dan's group completes attributes I'd like mining to have.
So for the first time, I need to tip my hat to Dan and kudos.
Daniel Larimer/Bitshares et. al claim some benefits which are explained at the following references:
https://steem.io/documentation/consensus/#mining-algorithm
https://steem.io/SteemWhitePaper.pdf#page=23
This is perhaps the first technical idea I've seen from the Dan Larimer group that I must frankly state is very good and I will likely adopt it for my own project. However, I would structure it differently to avoid the security caveats they noted. I would have the miner sign the hash(H, nonce), the use that as input to the PoW, e.g. PoW = hash(H,nonce,SIGN(hash(H, nonce))).
They think one of the benefits is that it will encourage the development of an ASIC which can perform faster elliptic curve signing and validation, and I presume this is so that they hope to be able to validate more transactions per second with more efficient hardware. But I don't claim that is a realistic expectation for a benefit, because well I expect to popularize unprofitable mining and thus the end of ASICs for mining forever.
The benefits I think are realistic are no botnets and no pools. Excellent. I had already eliminated farming out mining with another method (farming out can't be resistant to DDoS), and this idea from Dan's group completes attributes I'd like mining to have.
So for the first time, I need to tip my hat to Dan and kudos.
That sounds really strong!
And I found recently CIYAM speaking about a new 'proof-of-storage' here where I wonder if you could force now the miners also to have the entire blockchain stored locally for real:
https://bitcointalksearch.org/topic/m.15091671
Those mixed together is the holy grail ...
June 15, 2016, 01:36:25 PM
Steemit's PoW has a novel idea to incorporate signing into the PoW for the public key that will receive the coinbase block reward, so as to force the miner to have real-time access to the private key for each nonce attempted.
Daniel Larimer/Bitshares et. al claim some benefits which are explained at the following references:
https://steem.io/documentation/consensus/#mining-algorithm
https://steem.io/SteemWhitePaper.pdf#page=23
This is perhaps the first technical idea I've seen from the Dan Larimer group that I must frankly state is very good and I will likely adopt it for my own project. However, I would structure it differently to avoid the security caveats they noted. I would have the miner sign the hash(H, nonce), the use that as input to the PoW, e.g. PoW = hash(H,nonce,SIGN(hash(H, nonce))).
They think one of the benefits is that it will encourage the development of an ASIC which can perform faster elliptic curve signing and validation, and I presume this is so that they hope to be able to validate more transactions per second with more efficient hardware. But I don't claim that is a realistic expectation for a benefit, because well I expect to popularize unprofitable mining and thus the end of ASICs for mining forever.
The benefits I think are realistic are no botnets and no pools. Excellent. I had already eliminated farming out mining with another method (farming out can't be resistant to DDoS), and this idea from Dan's group completes attributes I'd like mining to have.
So for the first time, I need to tip my hat to Dan and kudos.
Daniel Larimer/Bitshares et. al claim some benefits which are explained at the following references:
https://steem.io/documentation/consensus/#mining-algorithm
https://steem.io/SteemWhitePaper.pdf#page=23
This is perhaps the first technical idea I've seen from the Dan Larimer group that I must frankly state is very good and I will likely adopt it for my own project. However, I would structure it differently to avoid the security caveats they noted. I would have the miner sign the hash(H, nonce), the use that as input to the PoW, e.g. PoW = hash(H,nonce,SIGN(hash(H, nonce))).
They think one of the benefits is that it will encourage the development of an ASIC which can perform faster elliptic curve signing and validation, and I presume this is so that they hope to be able to validate more transactions per second with more efficient hardware. But I don't claim that is a realistic expectation for a benefit, because well I expect to popularize unprofitable mining and thus the end of ASICs for mining forever.
The benefits I think are realistic are no botnets and no pools. Excellent. I had already eliminated farming out mining with another method (farming out can't be resistant to DDoS), and this idea from Dan's group completes attributes I'd like mining to have.
So for the first time, I need to tip my hat to Dan and kudos.
Jump to: