Topic: Proof of stake mining of bicoin - page 2. (Read 25676 times)

It's not FUD.  It's FACT.

Of course someone could control 51% of Bitcoin's hashpower or own 51% of all the NXT.  That isn't the crux of the argument.  The main contention with why Bitshares' DPoS is fundamentally flawed is that no one has to go out and acquire 51% of the hashpower or the stake.  All they have to do is convince stakeholders to vote them into power.  That is the added "social construct".

Would be nice to see a POS system introduced to Bitcoin but then again many wouldn't like it. I know if this would be in place would be nice to have this but how Bitconi was created I doubt very much they would add such things due to how big bitcoin is and how big it is becoming.  Who knows what the future hold maybe satoshi himself will come back and say ahh yes POW is fairly old so here is the new release with POS within it to generate some stakes.

Lol. How can you be sure that 51% of all bitcoin mining power isn't controlled by a single person, or that 51% of all NXT isn't held by or leased to a single person? Nice try at FUD, though.

Wrong.  This is a perfect example of why DPoS is broken.  We literally have a situation where the security of the network is at the mercy of stakeholders trusting delegates not to collude against them.  How can you be sure that other individuals aren't currently posing as multiple delegates?  YOU CAN'T!

With PoW or PoS, I can't control 51% of the hashpower or stake with the moving of my lips.

The fact that one individual was voted into multiple delegate positions is a breakdown of chain security.  Just because he didn't attack doesn't mean that he couldn't have or that someone in the future won't.  You can't say that "sfinder is a case of successful identification and removal out of malicious delegates" because YOU DON'T KNOW IF HE CONTROLS OTHER DELEGATES.  No one can be sure how many delegates are controlled by one person because DPOS IS FUNDAMENTALLY FLAWED.


After an attack is made, it is too late.  No one will care if the stakeholders remove their votes from the attacking delegates.  People will have lost money.  Who is going to be responsible for replacing investors' lost money?


This doesn't sound very "decentralized" to me.  It sounds like a system to bring about "approved validators" which is extremely similar to Ripple but more centralized.

This is perfect example of why DPOS works so well. We literally have a situation where the network autonomously removed a block producer the community identified as being malicious. With POW or POS there is no such possibility, if you have a malicious stakeholder/block producer you simply have to accept that they will stay so forever.

afaik sfinder didn't attack the network, he just ran several delegates while spreading bearish sentiment.  sfinder is a case of successful identification and removal out of malicious delegates.

You don't just have to lie, you have to convince stakeholders to vote for you enough times to win the majority of the delegate positions. The votes can be withdrawn and the delegates removed if an attack is made.  Any attack attempt may result in making BitShares immediately stronger as devs typically have to do some work in advance to get voted in.  They would have to do this 50+ times to have a chance at doing a double spend.  Basic block singing delegates are voted in more easily but they can still be kicked out and they will decrease in number as time moves forward.

Also there is an unique individual verification system being developed by http://followmyvote.com/ who are part of BitShares which could be used to ensure delegates are unique individuals and make it even stronger in the future.

throw your pentium 4 to garbage (bitcoin) and buy an i7 (nxt)

edit : nxt is core i5 and nem is core i7 ! now that's better !

You realize by saying this you are no longer a "cryptocurrency" and are now akin to Ripple which makes you trust that nodes (aka delegates) won't collude against you.


The cost of attacking Bitshares is extremely low.  It has already been done once.  How much did it cost "sfinder" to attack?  What makes you think it won't happen again?  What makes you think others delegates aren't currently deceiving people?  How much does it cost to mislead people through deception?


There is absolutely something wrong with introducing a social construct into chain security and then misleading people to think the system is "decentralized" and "Safer than a Swiss bank account."  Imo, PoW is evolving into a centralized system but it is still more decentralized than DPoS because it is mathematically provable through PoW that the chain is secured by a verifiable amount of hashpower.

If I want to attack Bitcoin's PoW algo, I have to acquire 51% of the hashpower.
If I want to attack NXT's PoS algo, I have to acquire 51% of the stake.
If I want to attack Bitshare's DPoS algo, ALL I have to do is LIE.

But isn't the attack a double spend? It can more than make up for lost revenue. If you run enough double spend scams for awhile you can always sell your position down and quietly exit.

Point taken there can be subcontracting. 

Mathematically proving delegates are different people isn't necessary.  Incentives work strongly against any attacker.  They would have to spend months developing (via subcontracting or doing it themselves) a bunch of features for BitShares to win voter approval to take multiple positions or they would have to buy a massive amount of BTS to be able to vote themselves in, only to then attack themselves by attacking the network.  The cost of attack is already very high.

There is nothing necessarily wrong with having a social construct be part of the mix.  Lets compare it to massive warehouses filled with asics run by a handful of mining pools.  Both methods are worthy experiments and no one knows what POW or DPOS will lead to in the future.

That's the most ridiculous thing to say in the internet age. Of course one person can do millions of procedural tasks at once. If you can't write software to do it you hire cheap tech support in Mexico or India and script them. Break them up into teams for each delegate pseudonym and offer compensation based on productivity.

This is a ridiculously naive argument with no appreciation for the reality of people's greed,
and 100% wrong.
 
The people that would abuse the system to gain multiple delegate spots
are NOT going to say "you know what, I'm being paid a fair amount already,
I don't want more than my fair share."

The more money there is to make, the MORE motivated everyone will be to cheat.
 

Epic fail is exactly right.  It's disingenuous, in every sense of the word, that the Bitshares' devs continue to lead people to believe that Bitshares is "decentralized" and I quote "Safer than a Swiss bank account!"  Obviously, they care more about getting investors' money than protecting it and creating an actual decentralized system.

The Bitshares' devs didn't find "sfinder" on their own accord.  Daniel Larimer (aka Bytemaster), the main Bitshares' dev, even voted him into multiple delegate positions.  It was some individuals in the Chinese community that found out about his multiple delegate positions.


I'm going to assume the first question is a joke.  I'm saying regardless of how many delegates you have 11, 101 or 1001, DPoS is fundamentally flawed because you geniuses forced a social construct onto chain security which opens you up to Sybil attacks.


Are you sure about that?  Prove it.  Oh wait... you can't because it is IMPOSSIBLE to in Bitshares' DPoS mechanism.


Yes, because we all know that all investors are diligent and have the experience necessary to ensure that they don't fall into a trap.  Also, we all know that scammers never, ever work together, right?  lol

How are you going to verify a delegate is doing his due diligence as a business if his progress can't be determined solely based on code commits?


I am "keeping it real".  Everyone involved in Bitshares that supports these claims of it being "decentralized" and "Safer than a Swiss bank account" is misleading those who don't know any better.  Bitshares destroyed PoS by adding delegates.  What is worse is that imo the real reason the Bitshares' devs added delegates was to steal NXT's PoS mechanism and try to make it look like they were "innovative" when really they just wanted to ride on NXT's success.  It is clear that Daniel knows that it is IMPOSSIBLE to ascertain the identities behind delegates, but they continue on this "decentralization" charade, imo purely to capitalize on the uninformed.

Now the Bitshares' cheerleaders will tell you, "What does it matter if the delegates are the same person?"  Surely, they jest.  The ENTIRE REASON behind ALL consensus mechanisms is to prevent SYBIL ATTACKS!  PoW does this with hashpower.  PoS does this with stake.  DPoS claims to be able to do this with "delegates", but this is impossible because multiple delegates can be controlled by one individual.  Some of you might say, well, with NXT's PoS, you can lease your forging power isn't that the same thing?  The difference is this.  NXT doesn't force the social construct onto the chain.  You DON'T have to lease your forging power to anyone!  There is NO reason to lease your forging power to anyone unless you want to receive the tx fees you earn on a more consistent basis at the cost of paying the forging pool operator a fee.  Leasing in NXT was created so you could lease your stake to YOURSELF and keep your main account offline.  Therefore allowing you to protect the network and not keep your main account open on your computer.

Anyone who supports Bitshares' DPoS consensus mechanism is either disingenuous, delusional or severely misinformed.


Nice try, but NXT's forging pools never got remotely close to GHash's hashpower concentration.  NXT also has way more than 101 forgers securing its chain.


I make it sound "easy", because it is "easy" to mislead and manipulate people who are uninformed.  Have you seen the scams that are being pulled over on people in the altcoin section?

This is the same problem being mulled over with my coin.  (Search RPCD ANN)

According to my analysis Bitcoin has incentives to centralize as market cap goes up, especially as long as block rewards are relatively high. You can't risk orphaned blocks mining on your own little home based node and p2p pool doesn't seem to be as efficient either.

Where with DPOS it seems all the centralization should occur early on and then as market cap grows, the largest reason for multiple delegates per person decreases because each can pay a fair amount of compensation. At that point shareholders (which POW miners are not) will be incentivized to keep the network decentralized because NETWORK SECURITY WILL DEMAND IT !

You make it sound so easy. Its interesting how preference sways our judgement. For instance, I can easily say the same for PoW coins:

All someone has to do is operate a pool which rewards block finders generously and has 0 fees, and the miners would eagerly pull in the Trojan horse

If anything, the Ghash incident showed that miners are not pro-active in protecting the network and it leaves PoW coins vulnerable. Will shareholders be more careful when it becomes larger and the voting is more distributed? Well, only time will tell.

Certainly more difficult than one individual or group running several pools and getting control over 51% hashrate. Imagine having around 10 delegates, imagine that the delegates have power proportional to the votes, imagine that 2-3 of the top voted delegates have control due to the proportionality factor - well, you get Bitcoin.

DPoS concedes that some degree of gravitating towards a few concentrated points is always going to happen, as was clear in the Ghash success, and later seen in NXT forging pools too. It attempts to regulate this behaviour by making it inbuilt in the protocol itself and trying to make sure that there is no more centralization than is necessary.

Sfinder was found and voted out successfully. 

It is almost impossible for full pay delegates to be the same person.  Unless they are superhuman and able to complete multiple jobs at the same time.  You can view the work delegates are doing.  They have to be transparent and productive or they are voted out.  It's simply not possible to all be the same person because 1 person cannot do 101 jobs.  And there are huge incentives to not do this.

Delegates are one of the great features of BitShares.  It's creating a thriving community and a whole new type of decentralised organisation.  Not to mention the unique identity verifying features being developed.   DPOS is an amazing, fascinating and powerful invention which is going to re-structure society itself.  Constructive criticism is always welcomed, but please keep it real.

Why is having multiple delegates belong to one person bad?
Or are you saying that 101 is indeed the magic number, and anything less is too centralized? What if we coalesced redundant delegates into single nodes?

Epic fail.  I knew it.  Not to mention that even if the delegates are known,
then they become subject to manipulation.

How did they uncover this "sfinder" character?