Made new video: "Solving Bitcoin's Centralization: NXT vs Clam"

[Marc De Mesel]

Enjoy  

"Solving Bitcoin's Centralization: NXT vs Clam"

https://www.youtube.com/watch?v=WMZqnbtzXOI


Summary: Bitcoin failed to remain decentralised due to it's proof of work design. Proof of stake is biggest innovation in crypto since invention of bitcoin. NXT and Clams are great POS coins but for very different reasons with very different risks. Now that altcoins are at an all time low these are great buys in my opinion.

[monsterer]

Proof of stake is biggest innovation in crypto since invention of bitcoin.

Proof of stake creates more problems than it solves. I'm quoting from another post of mine, but here are several attacks which are not present in POW:

I would add the following POS attacks to your list:

* Custodial stake

Exchanges and other large services which store user funds in their own wallets gather a very large stake, which often would give them majority power of POS block generation if they were to abuse it.

* Chain freeze

Once a majority stake holder becomes the dominant block producer, they can withhold all blocks forever, bringing the entire chain to a permanent halt, correctable only with a hard fork.

* Shorting attack

A whale takes out a large short of a POS coin at the same time he buys an equal portion of stake, such that his overall position is neutral.

He then uses his stake to double spend by creating blocks continuously (whenever he is permitted to do so) thereby driving the price of the currency down until he is ready to close his short in profit.

In addition, I would say the chief disadvantage of POS over POW is that the security model in POS is much weaker than POW; block generation probability/cost is a constant in the amount of stake you own, whereas in POW the cost of block generation is super linear in the number of blocks. This makes attacking a POS chain cost free under the shorting attack described above.

ref: https://bitcointalk.org/index.php?topic=1316024.msg13489124#msg13489124

[Marc De Mesel]

Proof of stake creates more problems than it solves.

I would add the following POS attacks to your list:

* Custodial stake

Exchanges and other large services which store user funds in their own wallets gather a very large stake, which often would give them majority power of POS block generation if they were to abuse it.

What we see in practice is that the dominant exchange for NXT (bter before, now poloniex) only have around 5% of the coins in their custody. They need 10 times more to do a 51% attack which is very unlikely to happen.

With Clam it is indeed the case that one successful business does have more than 51% of the coins in custody so it is more centralized than bitcoin. However it has the right design so that over time, if the coins is successful in gaining adoption, more and more people will validate transactions and it will become more decentralised, in contrast to bitcoin that over time becomes more and more centralized.


Also note that eventhough Clam is currently in constant 51% attack vulnerability the likelihood of a 51% attack is much lower than bitcoin as the custodian would harm himself the most of all by destroying this coin where his whole business depends on. Bitcoin however, due to it's high market cap and visibility, has many enemies that could easily invest 100 million in mining equipment or threaten a few mining operators to do a successful 51% attack.

* Chain freeze

Once a majority stake holder becomes the dominant block producer, they can withhold all blocks forever, bringing the entire chain to a permanent halt, correctable only with a hard fork.

True, as I also explain in my video, once you succeed in a 51% attack with a POS coin and own 51% of the coins, you are unbeatable and can attack the coin continuously without anyone able to stop you. This however I don't see as a big problem. It's inherent to high security, that once it does break, you're fucked. The problem is that bitcoin has way too low security. You only need, not min 51%, but max 5% of the coins in bitcoin (and liquidate them for mining equipment)  to do a successful 51% attack.

* Shorting attack

A whale takes out a large short of a POS coin at the same time he buys an equal portion of stake, such that his overall position is neutral.

He then uses his stake to double spend by creating blocks continuously (whenever he is permitted to do so) thereby driving the price of the currency down until he is ready to close his short in profit.

I can't follow your reasoning here but I do know that to short you need the coins as well. Who is going to lend you 5%, let alone 20% of the coins? Extremely unlikely.

In addition, I would say the chief disadvantage of POS over POW is that the security model in POS is much weaker than POW; block generation probability/cost is a constant in the amount of stake you own, whereas in POW the cost of block generation is super linear in the number of blocks. This makes attacking a POS chain cost free under the shorting attack described above.

Nice theory but in practice NXT, the first 100% proof of stake coin is rolling for 2 years now and a 51% attack has not even come close to success. In the meantime many other Proof of Work coins have been attacked successfully and even bitcoin is in dire straits.

[monsterer]

Also note that eventhough Clam is currently in constant 51% attack vulnerability the likelihood of a 51% attack is much lower than bitcoin as the custodian would harm himself the most of all by destroying this coin where his whole business depends on. Bitcoin however, due to it's high market cap and visibility, has many enemies that could easily invest 100 million in mining equipment or threaten a few mining operators to do a successful 51% attack.

Not so - an exchange does not depend on one coin to run it's business; they typically have 10s of coins open for trading. In addition, as I mention below, the shorting attack can make this cost free.

This however I don't see as a big problem. It's inherent to high security, that once it does break, you're fucked. The problem is that bitcoin has way too low security. You only need, not min 51%, but max 5% of the coins in bitcoin (and liquidate them for mining equipment)  to do a successful 51% attack.

In bitcoin you need a continuous investment of 25 BTC per block in order to attack the chain. That's $1,605,600 per day at current prices. And as I've already explained, the shorting attack in POS makes attacking a POS chain cost free.

I can't follow your reasoning here but I do know that to short you need the coins as well. Who is going to lend you 5%, let alone 20% of the coins? Extremely unlikely.

Yet entirely possible.

Nice theory but in practice NXT, the first 100% proof of stake coin is rolling for 2 years now and a 51% attack has not even come close to success. In the meantime many other Proof of Work coins have been attacked successfully and even bitcoin is in dire straits.

The absence of a thing does not prove anything. Theory matters. A lot.

[Thenoticer]

Talk is cheap.

Who will attack NXT? Please use these exploits to attack. Attacker gets to win money.


Win money attack NXT with actions not words.

[tokeweed]

Enjoy  

"Solving Bitcoin's Centralization: NXT vs Clam"

https://www.youtube.com/watch?v=WMZqnbtzXOI


Summary: Bitcoin failed to remain decentralised due to it's proof of work design. Proof of stake is biggest innovation in crypto since invention of bitcoin. NXT and Clams are great POS coins but for very different reasons with very different risks. Now that altcoins are at an all time low these are great buys in my opinion.

If so, then what's your opinion on PPC?  And will it traction in the future? 

[Marc De Mesel]

Enjoy  

"Solving Bitcoin's Centralization: NXT vs Clam"

https://www.youtube.com/watch?v=WMZqnbtzXOI


Summary: Bitcoin failed to remain decentralised due to it's proof of work design. Proof of stake is biggest innovation in crypto since invention of bitcoin. NXT and Clams are great POS coins but for very different reasons with very different risks. Now that altcoins are at an all time low these are great buys in my opinion.

If so, then what's your opinion on PPC?  And will it traction in the future? 

Peercoin, although the first coin with proof of stake, looks very much like a copycoin when you download the client, which put me off to further investigate.


There are many POS coins. What I like to see is unique utility.

I really need nxt if I want to buy or launch an asset on it's decentralised exchange.

I really need clams to gamble on just-dice.


For what do I really need ppc that I cannot do with bitcoin?

[enet]

The issue of Proof-of-Stake is not 51% attack, but 51% control. In practise it means that is a quite centralised system. Even if 51% of the stakeholders have the interest of the system in mind, they can dictate the direction. Nxt original code and ideas are great. It is however not a very open system - the code is tightly controlled. A more distributed system could also ensure that the project follows some guidance by users and features and services they might want. Its very hard to find the balance between a coherent vision and input of many people (design by committee problem). The last 12 months have been pretty quiet. It would be good to see more R&D in this direction and more new systems.

[smooth]

What we see in practice is that the dominant exchange for NXT (bter before, now poloniex) only have around 5% of the coins in their custody. They need 10 times more to do a 51% attack which is very unlikely to happen.

No, you only need 51% of the coins that are actively staking, not 51% of the total supply. This is why exchange hacks have been a huge deal with PoS coins even when they don't approach 50% of the total supply.

In the case of Nxt (and most coins today that aren't really used much) the bulk of the supply is likely held relatively few people, each having large amounts, which motivates staking (and also makes them feel personally invested in the coin for a significant portion of their net worth, which means they will actively defend against attacks, while someone with pocket change would not particularly give a shit). If I own 300 EUR worth of coin out of a supply worth billions or even trillions of EUR, I'm just not ever going to bother staking it.

Some coins try to get around this with delegated staking but the problem there is further centralization to one or a few stake pools.

CLAM does at least have an incentive to actively stake since it uses proof-of-working-stake (no coin-age component). Still many small holders of CLAM don't bother to stake in practice.

Talk is cheap.

Who will attack NXT? Please use these exploits to attack. Attacker gets to win money.

Win money attack NXT with actions not words.

It may be that Nxt is in fact 100% secure because >50% of the stake has always been held by a few people such as Marc De Mesel who will defend it against attacks. That doesn't make it decentralized nor does it make the method secure just because Nxt in particular might be.

[gaba]

If I own 300 EUR worth of coin out of a supply worth billions or even trillions of EUR, I'm just not ever going to bother staking it.

In NXT you don't need to stake with active connection or online machine. You can lease your account to other address, maybe to some developers or whoever. At the same time you can spend your NXT when and wherever you like.

Just imagine how will that help CLAM. I know you are running CLAM pool but I don't trust you. In NXT you don't need that trust. My NXT is still mine.

[smooth]

If I own 300 EUR worth of coin out of a supply worth billions or even trillions of EUR, I'm just not ever going to bother staking it.

In NXT you don't need to stake with active connection or online machine. You can lease your account to other address, maybe to some developers or whoever. At the same time you can spend your NXT when and wherever you like.

See above, search for delegated staking.

Concentrating stake like that does not make it more secure. It just creates a central point of control or potential compromise (custodial risk with respect to staking as described by monsterer is the same here, in fact worse because people are more likely to concentrate stake with centralized entities if their coins can't also be stolen).

[enet]

Smooth - good points. The fee problem hasn't been solved by any system AFAICS. Transaction validation should be fairly compensated and to such a degree that it is roughly profitable (enough incentives so that supply can serve demand).

In terms of fairness and distribution. Imagine a coin called AliceBobCoin. Alice holds 51% of all coins and Bob holds 49% of all coins. Is it decentralised? Imagine a PoS coin called EveryoneCoin with 1 billion users where everyone has the same share 10^-9. Interestingly fairness is probably not even the best factor. It would be desirable for the system to compensate productive behaviour. Instead of everyone getting money for free or by lottery, its better to agree on some other method which rewards behaviour. In just the same way a company wants shareholders to be productive members and aligning with the interest of the corporation.

[DecentralizeEconomics]

It may be that Nxt is in fact 100% secure because >50% of the stake has always been held by a few people such as Marc De Mesel who will defend it against attacks. That doesn't make it decentralized nor does it make the method secure just because Nxt in particular might be.

I guess you have to ask yourself, "How much does decentralization matter if the currency holders are the ones that secure the network?"

[smooth]

Smooth - good points. The fee problem hasn't been solved by any system AFAICS. Transaction validation should be fairly compensated and to such a degree that it is roughly profitable (enough incentives so that supply can serve demand).

In terms of fairness and distribution. Imagine a coin called AliceBobCoin. Alice holds 51% of all coins and Bob holds 49% of all coins. Is it decentralised?

Certainly not. It is entirely controlled by Alice. Alice might as well (and it would be vastly more efficient to) just keep the coin balances on MySQL.

But fairness is not required. A million holders where one person holds 1000 times more than each other person is still pretty decentralized.

(I assume we are talking about PoS here and ignoring other issues with PoS.)

[smooth]

It may be that Nxt is in fact 100% secure because >50% of the stake has always been held by a few people such as Marc De Mesel who will defend it against attacks. That doesn't make it decentralized nor does it make the method secure just because Nxt in particular might be.

I guess you have to ask yourself, "How much does decentralization matter if the currency holders are the ones that secure the network?"

It certainly does. The point of cryptocurrency is not Moon. The point is to have a system of enforced rules where all participants are protected not only from outside attackers but from each other.

[TPTB_need_war]

PoS(hit) can never be secure, because if it has a functioning markets (which it must in order to be widely adopted and liquid), then one can borrow stake, attack the coin (which requires much less than 51% to for example delay transactions by some N blocks where N is a function of percentage of coin supply held), and then pay back the borrowed coin with cheaply bought coin as the price collapses due to attacks. You could simultaneously short it (i.e. which you did when you borrowed the coins, but sell some for fiat before you attack) for profits. Alternatively borrow fiat (or other cryptocoin), buy stake and short to profit and pay back loan. Also PoS can't distribute new coins, thus eventually the coin supply shrinks asymptotically to 0.

With PoW, your borrowed mining hashrate would eventually reach end of contract and the coin would repair itself. And you'd need much closer to 51% to do damage. You would hope to be able to purchase the coin at cheap prices, wait for it to rise back up and then sell it for fiat to pay back your loan. Much less plausible.

However if you are up against the corrupt State that charges cost of PoW mining to the collective, then we're screwed with profitable PoW also, except I have the idea to use the unprofitable PoW of every person's computer in the world (with latency preventing them from farming out to ASIC), which seems might be even too much of an expense for China to hide the subsidization of.

[CoinHoarder]

Also PoS can't distribute new coins, thus eventually the coin supply shrinks asymptotically to 0.

You are wrong here. There are PoS variants that distribute new coins. You guys all too often stereotype all PoS coins while forgetting there are many different variants with different pros and cons. Any argument against a PoS implementation should at least mention what version of PoS you are referencing.

PoW has many pros and cons that PoS coins don't have, and visa versa. I don't understand why everyone feels the need to push one or the other as superior, when they both have useful features that the other doesn't.

[CoinHoarder]

It certainly does. The point of cryptocurrency is not Moon.

I find it amusing when people try to force their vision as to what cryptocurrencies and blockchain technology should be onto others.

[smooth]

Also PoS can't distribute new coins, thus eventually the coin supply shrinks asymptotically to 0.

You are wrong here. There are PoS variants that distribute new coins.

To whom?

[TPTB_need_war]

Also PoS can't distribute new coins, thus eventually the coin supply shrinks asymptotically to 0.

You are wrong here. There are PoS variants that distribute new coins.

No variants can. And the last time you debated me, I defeated you on every single point. Are we going to have to do it again?

I thought you wrote last time that you were going to be busy with your university semester beginning. I am ~51 (50.7), you are in your early 20s I presume. I have been a software developer since before you were born.