4 ms·

> Proof of stake is a scam. When I say that, I mean that proof of stake is (1) claimed to be a consensus system, and (2) constitutionally incapable of actually

by darawk 5y ago

> Proof of stake is a scam. When I say that, I mean that proof of stake is (1) claimed to be a consensus system, and (2) constitutionally incapable of actually producing a consensus. Ok. Go break one of the many existing systems that operates using proof of stake then. If you've done this, you should be leading your article with it. If you haven't, you shouldn't be speaking. Proof of stake is not some theoretical thing being proposed in the abstract. Many systems operate on it as we speak.

inter_netuser 5y ago
you didn't read the article to the end, did you? about 40% in: "Because of all the arguments above, we can safely conclude that this threat of an attacker building up a fork from arbitrarily long range is unfortunately fundamental, and in all non-degenerate implementations the issue is fatal to a proof of stake algorithm’s success in the proof of work security model. However, we can get around this fundamental barrier with a slight, but nevertheless fundamental, change in the security model." —Vitalik Buterin, saying the quiet part out loud Security model in PoS = trust the rich. Some like having masters, whatever floats your boat.
- darawk 5y ago
  > you didn't read the article to the end, did you? I skimmed it. It made no serious arguments. If it had a serious argument, it would have exploited one of the many existing proof of stake systems. > Security model in PoS = trust the rich. Some like having masters, whatever floats your boat. You mean...exactly like PoW mining?
  - inter_netuser 5y ago
    No, exactly the opposite of PoW mining. Miners do not set the rules, they are merely a service that provides immutability to a ledger, with a nuclear option that will bankrupt all the billions they have invested, should they misbehave. Large stakers can rent-seek and extract your wealth, PoS is the same system we have now, plus some code. You are quite literally being exploited right this minute, by the same methods outlined in the article.
    - darawk 5y ago
      > Miners do not set the rules, they are merely a service that provides immutability to a ledger, with a nuclear option that will bankrupt all the billions they have invested, should they misbehave. Stakers do not set the rules any more than do large mining pools. > Large stakers can rent-seek and extract your wealth, PoS is the same system we have now, plus some code. You know MEV is a thing in PoW too, right?
    - dathinab 5y ago
      Miners decide what they mine. Their contribution is what makes the currency work. They are some of the main profiteers from the currency. You can't do a fork or rule change without enough miners going along (I mean you can but it would not have much value). So in practice have all the governing/decision making power (as a group not a single person). Sure they might not come up with changes, but they do decide weather any change do take effect in the end.
    - reyqn 5y ago
      I'm not really knowledgeable about all this, but mining of PoW currencies right now seems to rely a lot on mining pools. Isn't there a risk that they are "the rich" and people trust them? What's the difference with PoS there?
- nootropicat 5y ago
  Observing what everyone else is using isn't 'trusting the rich'.
- dathinab 5y ago
  > Security model in PoS = trust the rich. Same for PoW. Crypto currency weather PoW or PoS boils down to "give the few rich all the power while giving the many less rich a illusion of security". In PoW it just slightly tweaks "richness of money" into "richness of computation resources (which you get through money)". This difference has complicated effects like: - benefits anyone with cheap electricity (i.e. either places with no environmental protection, government support in some way, or the few places with cheap clean power) - benefits anyone with good connections to chip factories - the investment needed for gaining power being less bound to the currency itself but computation power instead
hopevillian 5y ago
Ethereum is literally broken, look it up
- darawk 5y ago
  Sure, go steal all the money in curve. I'll wait.
  - hopevillian 5y ago
    yeah, save all that money in a monetary system that is controlled and literally changed on the whim of it's creators... sounds vary familiar to another monetary system we currently are on... oh, well here are some interesting links https://www.youtube.com/watch?v=P8LXLoTUJ5g https://www.youtube.com/watch?v=P8LXLoTUJ5g https://twitter.com/CryptoLanroc/status/1458144922867957761 https://twitter.com/CryptoLanroc/status/1458144922867957761 https://betterprogramming.pub/the-encyclopedia-of-smart-contract-attacks-vulnerabilities-dfc1129fdaac https://betterprogramming.pub/the-encyclopedia-of-smart-cont... https://twitter.com/SpillyGuy/status/1432370045758447625?s=20 https://twitter.com/SpillyGuy/status/1432370045758447625?s=2...
    - darawk 5y ago
      So by "broken" you mean you can't actually break anything. Got it. MEV is well known, and the system operates just fine in its presence.
      - light_cone 5y ago
        It's incredible how people who read a few articles think they have found the smoking gun, against a technology on which a lot of crypto researchers devote years of their time to make and analyze. They must think themselves are geniuses, and researchers are fools. Oh well, nothing new under the sun.
akrymski 5y ago
If you'd read the article till the end, perhaps you'd understand where the author is coming from: PoS systems aren't getting hacked today because they aren't truly decentralised. You can't have decentralisation and security with POS, you have to choose one of these. All the projects currently active have chosen security for obvious reasons - they control the majority of validators to make sure nobody steals, and are just fancy centralised mints.
- darawk 5y ago
  Except that they are though...You can go run a validator on Eth or Cardano right now.
  - akrymski 5y ago
    Sure, but if the majority of validators are actually run by the project owners it's effectively centralised. And it's easy to maintain this control if you have the majority of coins to stake. It's not Sybil resistant for this reason - all validators could be owned by one person and you would have no way of knowing.
    - dorgo 5y ago
      >all validators could be owned by one person and you would have no way of knowing. all bitcoin miners could be owned by one person and you would have no way of knowing.... >it's effectively centralised so your argument is not that it is not possible to have security AND decentralisation with POS, but that it currently is not the case, right?
      - akrymski 5y ago
        > all bitcoin miners could be owned by one person and you would have no way of knowing.... Sure, it's entirely possible that BTC is also centralised and controlled by wales. I was merely suggesting that the reason PoS systems haven't been hacked (much) yet is because the validators are controlled by project owners, so they are really centralised payment systems in disguise. There's a difference though: buying initial stake in PoS may be similar to buying an ASIC in PoW, but mining a chain has a real cost (electricity) in PoW. In PoS there's no cost to mining, so validators have an incentive to stake all possible forks. There's no way to have consensus on the correct chain, because real resources haven't gone into building one up.

by darawk 5y ago

inter_netuser 5y ago

you didn't read the article to the end, did you? about 40% in: "Because of all the arguments above, we can safely conclude that this threat of an attacker building up a fork from arbitrarily long range is unfortunately fundamental, and in all non-degenerate implementations the issue is fatal to a proof of stake algorithm’s success in the proof of work security model. However, we can get around this fundamental barrier with a slight, but nevertheless fundamental, change in the security model." —Vitalik Buterin, saying the quiet part out loud Security model in PoS = trust the rich. Some like having masters, whatever floats your boat.

darawk 5y ago
> you didn't read the article to the end, did you? I skimmed it. It made no serious arguments. If it had a serious argument, it would have exploited one of the many existing proof of stake systems. > Security model in PoS = trust the rich. Some like having masters, whatever floats your boat. You mean...exactly like PoW mining?
- inter_netuser 5y ago
  No, exactly the opposite of PoW mining. Miners do not set the rules, they are merely a service that provides immutability to a ledger, with a nuclear option that will bankrupt all the billions they have invested, should they misbehave. Large stakers can rent-seek and extract your wealth, PoS is the same system we have now, plus some code. You are quite literally being exploited right this minute, by the same methods outlined in the article.
  - darawk 5y ago
    > Miners do not set the rules, they are merely a service that provides immutability to a ledger, with a nuclear option that will bankrupt all the billions they have invested, should they misbehave. Stakers do not set the rules any more than do large mining pools. > Large stakers can rent-seek and extract your wealth, PoS is the same system we have now, plus some code. You know MEV is a thing in PoW too, right?
  - dathinab 5y ago
    Miners decide what they mine. Their contribution is what makes the currency work. They are some of the main profiteers from the currency. You can't do a fork or rule change without enough miners going along (I mean you can but it would not have much value). So in practice have all the governing/decision making power (as a group not a single person). Sure they might not come up with changes, but they do decide weather any change do take effect in the end.
  - reyqn 5y ago
    I'm not really knowledgeable about all this, but mining of PoW currencies right now seems to rely a lot on mining pools. Isn't there a risk that they are "the rich" and people trust them? What's the difference with PoS there?
nootropicat 5y ago
Observing what everyone else is using isn't 'trusting the rich'.
dathinab 5y ago
> Security model in PoS = trust the rich. Same for PoW. Crypto currency weather PoW or PoS boils down to "give the few rich all the power while giving the many less rich a illusion of security". In PoW it just slightly tweaks "richness of money" into "richness of computation resources (which you get through money)". This difference has complicated effects like: - benefits anyone with cheap electricity (i.e. either places with no environmental protection, government support in some way, or the few places with cheap clean power) - benefits anyone with good connections to chip factories - the investment needed for gaining power being less bound to the currency itself but computation power instead

hopevillian 5y ago

Ethereum is literally broken, look it up

darawk 5y ago
Sure, go steal all the money in curve. I'll wait.
- hopevillian 5y ago
  yeah, save all that money in a monetary system that is controlled and literally changed on the whim of it's creators... sounds vary familiar to another monetary system we currently are on... oh, well here are some interesting links https://www.youtube.com/watch?v=P8LXLoTUJ5g https://www.youtube.com/watch?v=P8LXLoTUJ5g https://twitter.com/CryptoLanroc/status/1458144922867957761 https://twitter.com/CryptoLanroc/status/1458144922867957761 https://betterprogramming.pub/the-encyclopedia-of-smart-contract-attacks-vulnerabilities-dfc1129fdaac https://betterprogramming.pub/the-encyclopedia-of-smart-cont... https://twitter.com/SpillyGuy/status/1432370045758447625?s=20 https://twitter.com/SpillyGuy/status/1432370045758447625?s=2...
  - darawk 5y ago
    So by "broken" you mean you can't actually break anything. Got it. MEV is well known, and the system operates just fine in its presence.
    - light_cone 5y ago
      It's incredible how people who read a few articles think they have found the smoking gun, against a technology on which a lot of crypto researchers devote years of their time to make and analyze. They must think themselves are geniuses, and researchers are fools. Oh well, nothing new under the sun.

akrymski 5y ago

If you'd read the article till the end, perhaps you'd understand where the author is coming from: PoS systems aren't getting hacked today because they aren't truly decentralised. You can't have decentralisation and security with POS, you have to choose one of these. All the projects currently active have chosen security for obvious reasons - they control the majority of validators to make sure nobody steals, and are just fancy centralised mints.

darawk 5y ago
Except that they are though...You can go run a validator on Eth or Cardano right now.
- akrymski 5y ago
  Sure, but if the majority of validators are actually run by the project owners it's effectively centralised. And it's easy to maintain this control if you have the majority of coins to stake. It's not Sybil resistant for this reason - all validators could be owned by one person and you would have no way of knowing.
  - dorgo 5y ago
    >all validators could be owned by one person and you would have no way of knowing. all bitcoin miners could be owned by one person and you would have no way of knowing.... >it's effectively centralised so your argument is not that it is not possible to have security AND decentralisation with POS, but that it currently is not the case, right?
    - akrymski 5y ago
      > all bitcoin miners could be owned by one person and you would have no way of knowing.... Sure, it's entirely possible that BTC is also centralised and controlled by wales. I was merely suggesting that the reason PoS systems haven't been hacked (much) yet is because the validators are controlled by project owners, so they are really centralised payment systems in disguise. There's a difference though: buying initial stake in PoS may be similar to buying an ASIC in PoW, but mining a chain has a real cost (electricity) in PoW. In PoS there's no cost to mining, so validators have an incentive to stake all possible forks. There's no way to have consensus on the correct chain, because real resources haven't gone into building one up.