The N@S paper from Kushti is important, though, this attack vector is always brought out whenever Nxt (and POS in general) is discussed.
Right now, the paper takes a few improtant steps in demonstrating that a nothing-at-stake attack is almost ceratinly dooned to fail.
And Hi! to the Coinomat dude!
Topic: Something at stake - proof of stake alternative (Read 6497 times)
That's very uncharitable at Christmas! You can spam when the POI whitepaper is finally released 
il hold you to that..
im all for spreading the word on that paper though, i just could help myself ha. good that theres finally something solid out to tackle the pos fud. i see there is virtually zero arguments against it lol
That's very uncharitable at Christmas! You can spam when the POI whitepaper is finally released
@sasha, rigorous research in the the 'Nothing at Stake' assertions has been published >>> https://bitcointalksearch.org/topic/nothing-at-stake-long-range-attack-on-proof-of-stake-consensus-research-897488
Please comment in the thread >>> https://bitcointalksearch.org/topic/nothing-at-stake-long-range-attack-on-proof-of-stake-consensus-research-897488. It would be good if we could get heads together from all aspects of cyrpto in the same place to discuss.
Please comment in the thread >>> https://bitcointalksearch.org/topic/nothing-at-stake-long-range-attack-on-proof-of-stake-consensus-research-897488. It would be good if we could get heads together from all aspects of cyrpto in the same place to discuss.
Quote
why not bang the drum a little quieter Daedelus? Been away a day or so and catching up... you're in almost every thread I've looked at. Maybe it is our overlapping interests but sheesh!
i couldnt resist..
@sasha, rigorous research in the the 'Nothing at Stake' assertions has been published >>> https://bitcointalksearch.org/topic/nothing-at-stake-long-range-attack-on-proof-of-stake-consensus-research-897488
Please comment in the thread >>> https://bitcointalksearch.org/topic/nothing-at-stake-long-range-attack-on-proof-of-stake-consensus-research-897488. It would be good if we could get heads together from all aspects of cyrpto in the same place to discuss.
Thank you, I'm watching this, I'm "coinomat" at nxtforum Please comment in the thread >>> https://bitcointalksearch.org/topic/nothing-at-stake-long-range-attack-on-proof-of-stake-consensus-research-897488. It would be good if we could get heads together from all aspects of cyrpto in the same place to discuss.
I never knew!
@sasha, rigorous research in the the 'Nothing at Stake' assertions has been published >>> https://bitcointalksearch.org/topic/nothing-at-stake-long-range-attack-on-proof-of-stake-consensus-research-897488
Please comment in the thread >>> https://bitcointalksearch.org/topic/nothing-at-stake-long-range-attack-on-proof-of-stake-consensus-research-897488. It would be good if we could get heads together from all aspects of cyrpto in the same place to discuss.
Thank you, I'm watching this, I'm "coinomat" at nxtforum Please comment in the thread >>> https://bitcointalksearch.org/topic/nothing-at-stake-long-range-attack-on-proof-of-stake-consensus-research-897488. It would be good if we could get heads together from all aspects of cyrpto in the same place to discuss.
@sasha, rigorous research in the the 'Nothing at Stake' assertions has been published >>> https://bitcointalksearch.org/topic/nothing-at-stake-long-range-attack-on-proof-of-stake-consensus-research-897488
Please comment in the thread >>> https://bitcointalksearch.org/topic/nothing-at-stake-long-range-attack-on-proof-of-stake-consensus-research-897488. It would be good if we could get heads together from all aspects of cyrpto in the same place to discuss.
Please comment in the thread >>> https://bitcointalksearch.org/topic/nothing-at-stake-long-range-attack-on-proof-of-stake-consensus-research-897488. It would be good if we could get heads together from all aspects of cyrpto in the same place to discuss.
This approach with prepayment, bet or whatever you call it is getting hot now,
another project based on this http://tendermint.com/
another project based on this http://tendermint.com/
No that is not subjective. Even if ghash had >50% and even if they misbehaved, the longest chain would still be objective and unambiguous.
We are back to what works and is useful, but that's a different question.
I'm not a fan of PoW as it exists in Bitcoin with extremely large pools, but I still assert that it has a stronger consensus model that is more robust with respect to trust. Again, that doesn't even imply it is necessarily useful at all.
We are back to what works and is useful, but that's a different question.
I'm not a fan of PoW as it exists in Bitcoin with extremely large pools, but I still assert that it has a stronger consensus model that is more robust with respect to trust. Again, that doesn't even imply it is necessarily useful at all.
Real world shows that POW is not objective, gigahash could have more than 50% of hashing power easily. So they choose to behave themselves not to hurt the network. It is very subjective.
"Any crypto" does not have to include social consensus, as Vitalik explained. PoW is "objective." This doesn't mean it always works perfectly -- if you are completely disconnected from the longest chain, you can't ever choose the longest chain. But in choosing between two chains, there is no social consensus involved. And in the first case (disconnected) social consensus won't help you, so it adds nothing.
But what you were saying about things that work is ultimately exactly what I was saying above. Trying to "fix" nothing-at-stake is dumb.
Actually I can't see anything wrong with checkpointing setup sort of which is implemented in NXT now or the one Vitalik wrote about recently. Such systems worked, work, and seem to be able to work in future. Field tests are successful, this is the most important.
Any crypto will be some cross between social consensus and "technical" consensus, so if technical consensus is not guaranteed by theory it doesn't mean that the system can't function.
Any crypto will be some cross between social consensus and "technical" consensus, so if technical consensus is not guaranteed by theory it doesn't mean that the system can't function.
"Any crypto" does not have to include social consensus, as Vitalik explained. PoW is "objective." This doesn't mean it always works perfectly -- if you are completely disconnected from the longest chain, you can't ever choose the longest chain. But in choosing between two chains, there is no social consensus involved. And in the first case (disconnected) social consensus won't help you, so it adds nothing.
But what you were saying about things that work is ultimately exactly what I was saying above. Trying to "fix" nothing-at-stake is dumb.
Real world shows that POW is not objective, gigahash could have more than 50% of hashing power easily. So they choose to behave themselves not to hurt the network. It is very subjective.
"Any crypto" does not have to include social consensus, as Vitalik explained. PoW is "objective." This doesn't mean it always works perfectly -- if you are completely disconnected from the longest chain, you can't ever choose the longest chain. But in choosing between two chains, there is no social consensus involved. And in the first case (disconnected) social consensus won't help you, so it adds nothing.
But what you were saying about things that work is ultimately exactly what I was saying above. Trying to "fix" nothing-at-stake is dumb.
Actually I can't see anything wrong with checkpointing setup sort of which is implemented in NXT now or the one Vitalik wrote about recently. Such systems worked, work, and seem to be able to work in future. Field tests are successful, this is the most important.
Any crypto will be some cross between social consensus and "technical" consensus, so if technical consensus is not guaranteed by theory it doesn't mean that the system can't function.
Any crypto will be some cross between social consensus and "technical" consensus, so if technical consensus is not guaranteed by theory it doesn't mean that the system can't function.
"Any crypto" does not have to include social consensus, as Vitalik explained. PoW is "objective." This doesn't mean it always works perfectly -- if you are completely disconnected from the longest chain, you can't ever choose the longest chain. But in choosing between two chains, there is no social consensus involved. And in the first case (disconnected) social consensus won't help you, so it adds nothing.
But what you were saying about things that work is ultimately exactly what I was saying above. Trying to "fix" nothing-at-stake is dumb.
Actually I can't see anything wrong with checkpointing setup sort of which is implemented in NXT now or the one Vitalik wrote about recently. Such systems worked, work, and seem to be able to work in future. Field tests are successful, this is the most important.
Any crypto will be some cross between social consensus and "technical" consensus, so if technical consensus is not guaranteed by theory it doesn't mean that the system can't function.
Any crypto will be some cross between social consensus and "technical" consensus, so if technical consensus is not guaranteed by theory it doesn't mean that the system can't function.
"Any crypto" does not have to include social consensus, as Vitalik explained. PoW is "objective." This doesn't mean it always works perfectly -- if you are completely disconnected from the longest chain, you can't ever choose the longest chain. But in choosing between two chains, there is no social consensus involved. And in the first case (disconnected) social consensus won't help you, so it adds nothing.
But what you were saying about things that work is ultimately exactly what I was saying above. Trying to "fix" nothing-at-stake is dumb.
Actually I can't see anything wrong with checkpointing setup sort of which is implemented in NXT now or the one Vitalik wrote about recently. Such systems worked, work, and seem to be able to work in future. Field tests are successful, this is the most important.
Any crypto will be some cross between social consensus and "technical" consensus, so if technical consensus is not guaranteed by theory it doesn't mean that the system can't function.
Any crypto will be some cross between social consensus and "technical" consensus, so if technical consensus is not guaranteed by theory it doesn't mean that the system can't function.
Quote
People who want to do useful work should stop trying to "fix" or "disprove" nothing-at-stake and accept it as a fact and go from there.
We don't even have a proper description of how N@S would work on a specific coin yet. All I hear is some bold statements "n@s is a fact". Please, you and I don't know the math behind it enough to make such claims.
Let's prove n@s first before saying it can't be disproved, allright?
I don't understand how derivatives apply in a PoS system. Having 'control' of paper money in a closed system like a brokerage is not the same as actually having that cryptographic currency to stake.
It matters because nothing at stake means you have nothing to lose. If you lose coins but gain at least the value of those coins elsewhere, then you still have nothing at stake.
Nothing at stake is fundamental and can't ever be fixed. That doesn't mean it isn't possible to build a functional or useful currency system around that fact, but it will have some properties that will always differ from proof-of-work (where an actual thing -- energy -- is consumed and can't be "unconsumed" elsewhere in an offsetting trade). People who want to do useful work should stop trying to "fix" or "disprove" nothing-at-stake and accept it as a fact and go from there.
Quote
Individuals 'leasing' stake to other individuals though sounds like a big problem to me. And any community that allows such a market to develop is probably making a big mistake since it could greatly lower the cost of an attack.
It can't be prevented unless the market is highly regulated. That's exactly what ArticMine was saying.
...
Do you have any arguments that actually don't also apply to bitcoin?
Do you have any arguments that actually don't also apply to bitcoin?
Yes. It can be summarized in one word: Derivatives.
The issue here is that stake does not equal exposure and it is very easy to have a large stake and zero or negative exposure. The reason this has not yet become an issue with POS coins is that none of them have evolved to the point where a significant derivatives market has developed, even to the degree that exists today with Bitcoin and Litecoin. It is today possible to control 10000 USD worth of XBT or 5000 USD worth of LTC with 500 USD margin. I leave it to the reader to determine how much EUR, or gold one can control with 500 USD in margin at a typical FX broker.
My take remains that POS can only work in a highly regulated market and even there, if the regulators are not on the ball we get the near financial disaster that occurred in 2008. After all the typical large wall street bank is ultimately governed by proof of stake. It is called one share = one vote.
Edit: Size of derivatives markets: http://www.bis.org/statistics/derstats.htm
I don't understand how derivatives apply in a PoS system. Having 'control' of paper money in a closed system like a brokerage is not the same as actually having that cryptographic currency to stake.
Individuals 'leasing' stake to other individuals though sounds like a big problem to me. And any community that allows such a market to develop is probably making a big mistake since it could greatly lower the cost of an attack.
My take remains that POS can only work in a highly regulated market and even there, if the regulators are not on the ball we get the near financial disaster that occurred in 2008. After all the typical large wall street bank is ultimately governed by proof of stake. It is called one share = one vote.
To me it seems to work right here right now, in an unregulated market. Would you say that Nxt, peercoin, etc. do not "work", or would you say that they are situated in a regulated market?
The whole point of PoS in crypto is that the wallstreet PoS in the sense of 1 share = 1 vote is centralized. Who guarantees that my shares are not diluted (ok, some cryptos sadly inflate supply as well)? If the management does something illegal, the company that I invested in could be taken down by the government, etc.
Yes. It can be summarized in one word: Derivatives.
Yes, as the arguments against POS are accepted as defeated, we must invent newer and even less relevant but more convoluted arguments to protect proof of waste.
hahahahaha, proof of work could really be called proof of waste.
Just so you don't miss it benjamin_bit...
https://nxtforum.org/consensus-research/multibranch-forging-approach/
First findings of the POS research group.
https://nxtforum.org/consensus-research/multibranch-forging-approach/
First findings of the POS research group.
https://bitcointalksearch.org/topic/nothing-at-stake-robust-pure-proof-of-stake-871576
Take a look at this thread, which covers the same topic.
Essentially to resolve nothing at stake, you need to prohibit all forms of txns among POS miners.
You do not need checkpoints.
I discuss the theory behind this. It sounds very restrictive, but there are some ways of loosening it up.
Take a look at this thread, which covers the same topic.
Essentially to resolve nothing at stake, you need to prohibit all forms of txns among POS miners.
You do not need checkpoints.
I discuss the theory behind this. It sounds very restrictive, but there are some ways of loosening it up.
https://blog.ethereum.org/2014/11/25/proof-stake-learned-love-weak-subjectivity/
Probably you've seen this, the latest on Stake from Vitalik Buterin. The proposed solution is check-pointing actually.
He mentioned something similar to what is proposed here, but discards it since he considers a security deposit, not a bet.
Probably you've seen this, the latest on Stake from Vitalik Buterin. The proposed solution is check-pointing actually.
He mentioned something similar to what is proposed here, but discards it since he considers a security deposit, not a bet.
Jump to: