Topic: It's about time to turn off PoW mining - page 16. (Read 39732 times)

Near the pole. But warm water are always needed 24/7 regardless of location

85% ?? Where did you get that number from?

http://www.cryptocoincharts.info/pair/bitusd/usd/bter/alltime

If you use the feature on the above website to hide the price spikes sufficiently, which are caused by low market depth and liquidity on BTER, you will see bitUSD has averaged 99.5% of USD's value. Furthermore, the centralized exchanges won't track the price as accurately as the decentralized market in BTSX, as most of the market depth and liquidity is on the decentralized exchange in BTSX.

However, this has nothing to do with PoW, PoS, or DPOS.

yes I know, it is marketing talk that is the point. It is sold as one thing but it is another.  I know that it has "pegs" and it will fail as soon as the base price of bitsharesX crashes atm it is only about 85% I guess that is a stable peg.

It's time we use the term Watts/Verification ?

Thanks for the reply! The quality of the discussion is getting better!

Right, it doesn't matter as long as we compare all systems as if they were not using inflation to pay for security. Bitcoin POW does have huge inflation atm to pay miners / pay for security.

Interesting! But then you would have the problem of actually nothing at stake (an attacker would just have to create millions of full nodes) plus own stake. In the end you have NXT like POS with the additional requirement to create a ton of full nodes to attack the network. It could be said that an attacker might be better at creating full nodes at scale than the average stakeholder which would make the system more easy to attack.

An attacker would have to trick stakeholder into thinking that his delegates are good delegates. The DPOS implementation of BTSX is in its early days. Today almost all delegates are some reputable forum members. Later the idea is that big companies or trustworthy public individuals from the crypto (currency) world would provide delegates: Exchanges, payment processors, DAC developement companies, trustworthy stakeholders; basically everyone that has an interest in a working BTSX system provides a delegate. Those entities then would have a lot more to loose (credibility of an exchange for example) than they could gain from a 51% attack which might be reversed anyway by a fork.
Regarding your attack: Stakeholders would have to be convinced that the random flood of delegates the attack set up are more trustworthy then the current delegates. That might work with 1 or even a few but not with 52.

Now if that happens there could be a fork that reverses the double spent / reversed transactions. Like you said correctly this might harm others who made transactions between the point in time when the double spend happened and when transactions are halted to fork. Those ppl could be compensated through a fund. The advantage would be that the incentive for a double spend attack would decrease to zero when the attacker knows that there is such a fund and a fork would reverse his double spend anyway. How could such a fund be funded (work only work if BTSX grows and tx volume picks up; but gives you an idea what delegates could be good for apart from tx processing): Buy a delegate or several delegates who commit to paying parts of their delegate pay into such a fund. Those delegates then campaign among Stakeholders with this fund.

As for forking to protect a network in general: Let's assume someone has acquired 51% of all hashing power for Bitcoin's sha256 mining then Bitcoin would have to find and implement a new proof of work algorithm in order to fork and get rid of the 51% control of a single party.
When someone has acquired 51% of stake in a POS coin you can simply fork and not honor this 51% and let the market decide which network is considered more trustworthy (probably the one without the 51% control....).

See above. And I would also say that a mix of anonymous and publicly known delegates would be most resilient.

Overall I do not propose that DPOS / POS has no disadvantages. I am pretty certain though that the security per cost you achieve with POS/DPOS is a lot higher than with POW. In the end security and decentralization costs and someone (coin holders or users) has to pay those costs.

Agreed, Inflation is a non-sequitor to this discussion.



What I am proposing is allowing full nodes to automatically become "delegates"  if they are on a unique IP and have a minimum stake of lets say 0.1BTC(But this could be adjusted as needed). The reasons for this have been explained in detail in the last 2 pages.

You are correct that with BTSX DPoS there is the cost of transferring part of ones stake to the intended candidate, but is this really a cost when all an attacker is doing is transferring his part of his stake from a shill voter to one of many shill delegates(to himself)? An attack with DPoS would simply involve flooding the market with many potential delegates, promoting those delegates to get enough other votes , and than using your shill accounts to tip the race in your favor for all candidates you control.  

The solution of forking after the attack has happened isn't satisfactory either because any hard fork because of an attack does huge damage to everyone. Forking is possible with any open source coin and what we are talking about is methods of securing a protocol so a hard fork after an attack isn't needed in the first place.

Please explain to me how this attack could be mitigated without decloaking the identity of all the potential delegates and thus bringing into the conversation a whole new can of worms.

Why the heck should POS be correlated with inflation??

This is not how it works. But kind of similar, see http://wiki.bitshares.org/index.php/DPOS#Voting_Algorithm

You are making the assumption that someone has a huge stake. If someone has more than 51% of stake in a POS system it is under the control of the person. But there would be a final solution to this: Fork and not honor this 51% stakeholder.

It seems to me like you tread DPOS and your proposed PoN in one sentence. Also it would make sense to differenciate between delegates and nodes. I have to admit that I didnt get your argument really.  

I hope you not say bitcoin should use PoS, because it will make inflation & the price will down

You are partially correct. A large stake holder could attempt to gain more delegates(nodes) by simply splitting up his coins (stake) amongst many nodes.

This problem is worse within DPoS as delegates can be anonymous(and if they aren't than that presents a whole other set of problems). If you have a large stake in DPoS of 60% you don't have to use all your voting power on only one of the 100 delegates but can create multiple wallets with multiple votes grabbing many different delegates positions. In fact you could potentially gain 80-90% of all delegate positions with a stake of 30-60% because all you need to do is flood the voting pool with potential delegates that you control and tip the odds in your favor for the close races.

With PoN there are much larger inherent costs to this task as an attacker needs to pay for or secure a large set of servers with different Ips and have a sizable amount of stake to spread amongst those servers to meet the minimum thresh-hold. With DPoS there is a limited amount of delegate slots and an attacker could flood the market with potential delegates, and with this there wouldn't be any limits to the amount of nodes(and the starting amount of nodes would be much higher) so an attack would be much more difficult. The incentives wouldn't be there to profit as transaction fees won't be motivation enough and the only objective would be to attack the network (and oneself) which would be held in check by the PoW consensus method. So this form of attack would cost a high amount in order to get a free coffee before the first PoW confirmation came in?

Nxt is much different than PoN because there is no need to invest in all the server and IP resources and your vote is only equal to the amount of stake you have directly. If Bitcoin had a minimum stake amount of 0.1 BTC per node to become a delegate than getting getting enough nodes , managing them, and securing them would be cumbersome and expensive.

"Your own Ft. Knox" is an a bit far fetched analogy. It is marketing talk. This should not be confused with what makes it work and what the functional output is. Functionally it can be best described as a prediction market which results in a crypto currency like token that is pegged to a certain real world asset  (e.g. the dollar or the gold measure in the price of 1 fine ounce of gold for example).
But let's not take the thread too much off topic. Blockchain security is a wide enough topic already. [Feel free to open a new thread and I will comment on it then! Just send me a pm so I know you opened it]

That is a team of developers, marketers and partners who work on the infrastructure of all the bitshares DACs. I3 or 3I means Invictus Innovations Incorporated. http://bitshares.org/community/team/

Go to the bitsharesX homepage, what slogan do you find? "Your own Ft. Knox" Yes it is totaly about ft. knox, that is a great example because we all know that not all the gold that is in ft. knox is there, and in bitsharesX there is ZERO gold behind bitGOLD and ZERO USD behind bitUSD, you can only cash out in bitsharesX. The USA would rather goverments get their payout in USD than Gold but they are not accepting that.

Right. Just as much as the biggest mining farms could.

That is correct: Every delegate would have an equal vote. But there would be more delegates than individuals who control those delegates. I think my previous post has shown why.

It is in now way about Ft. Know. No Gold or anything is deposited. It is only a prediction market that tracks prices of real world assets and makes the result of the prediction market sendable and divisible like you any crypto currency.

This is a great post, a good framework for change and a step in the right direction if implemented.

Who are the 3i team?

BTSX gives you the ilusion that it could be pegged to any asset? NO WAY, IT is much better than that, "IT IS YOUR OWN FT. KNOX"!

That says it about all, Ft. Knox where countries that depost there gold, and have no permission to inspect if their gold is really there.

LOL, thanks for the good laugh.... reminds me of this dialogue from Monty Python .... https://www.youtube.com/watch?v=fxGqcCeV3qk

There are some purposes in life where we need complex politics, but with currency I believe it is obsolete as the politics is represented by the code directly that individuals voluntarily agree to when they use it without creating the unnecessary need for abstractions of a personality contest.

Thanks for your criticism, keep it coming .....

I have explained that I do indeed see value in PoW for securing bitcoin so am not suggesting one migrate completely over to a PoS variant but possibly solve some of bitcoins current shortcomings with a hybrid approach of consensus. Suggesting we remove PoW from Bitcoin I believe is ultimately futile regardless, as everyone admits there is too much momentum and vested interests in PoW.

In this proposal of a hybrid confirmation system where some confirmations come from delegates(nodes consisting of miners and users) and some confirmations came from PoW (kept the same with average 10 min confirmations) a 51% would be more difficult to pull off because having a majority of the hashing power alone wouldn't be enough and one would have to control both a large stake and a large amount of nodes on separate IP's ranges. The technical aspects of accomplishing this hybrid approach would need to be teased out and tested heavily but there are several methods for accomplishing this.

The direct differences of this "Proof of Node" scheme and Nxt are significant. With Nxt your vote is based upon your stake where 10-15 large whales at the moment control a majority of the funds and can mount a 51% attack. With this Proof of Node every delegate has an equal vote despite the amount of stake like in DPoS and using simple techniques like having a minimum stake requirement within the node and requiring each node to exist on a different IP range would mitigate any risks of a botnet of nodes being created. So what I'm suggesting is just a simpler version of DPoS that accomplishes the same goals but more securely and with less costs.

There is a difference between a (full) node and a block producer (miner with POW, forger with NXT POS, delegate with DPOS). Full nodes can not prevent a 51% attack.
With your proposal above you would have the problem that the biggest stakeholders would distribute their stake among many accounts so that all accounts are above the defined threshold you mentioned. This in the end would equal NXT proof of stake. Not bad but no problem solved. 

I think you are really overstating how much development is being encouraged and done by the ongoing voting process. The development is primarily being done by the initial investment of  5,624.7 BTC investment (http://www1.agsexplorer.com/ ) . This fundraising retroactively paid for all of the development costs by the creators and has been blowing through 50k a week where it appears they will be running out of this capital soon (possibly in 26 weeks).  
Bitcoin has already grown to a size where a mixture of prestige , challenge, donations, and vested interests will contribute as many large open source projects, and I don't see the same happening with BTSX. Do you really believe the paltry sums from transaction revenue spread across 100 delegates(many of which won't contribute towards dev)  will support dev work in BTSX once the initial large investment runs dry?

I think you are stuck on discussing the benefits of DPoS vs PoW or DPoS vs current iterations of PoS.  I already agreed with you there are some benefits (and drawbacks as well).

The discussion has now turned to if we wanted to integrate a change to bitcoin than lets make sure the change is the right one. I see DPoS as an unnecessary process to accomplish the goal of adding security from the inherent weaknesses in PoS v1 coins. There are different ways of securing the network and increasing decentralization as I just proposed with a "Proof of Node" concept that doesn't involve the wasted time and complexity of a popularity contest.

Having "Full nodes" automatically become delegates where a minimum stake needed to prevent a botnet of delegate nodes forming would do the trick and insure there was a much larger pool of delegates at less of a cost to secure (less barriers to entry and thus more competition to become a delegate).

Are you suggesting there aren't the right incentives within this scheme vs DPoS? Many people already run full nodes who aren't miners without getting paid to do so, so I don't see how bitcoin couldn't dramatically increase from the current 20k-50k+ nodes(Hard to say exactly how many nodes exist as many come on and off line) to something much higher.