POW vs. DPOS - the worst miners / delegates could do to the network

[bytemaster]

back to topic:
https://bitcointalk.org/index.php?topic=664146.msg7736134#msg7736134

In fact, there is no poor user participation in NXT. Analysis of the blockchain reveals that the total number of accounts generating new blocks in NXT is between 300 and 350. Which is better than 100 delegates in BitShares.

Not if 300 of those accounts amount to 1% of the network.   Accounts are meaningless, only shares matter.

[xeroc]

back to topic:
https://bitcointalk.org/index.php?topic=664146.msg7736134#msg7736134

In fact, there is no poor user participation in NXT. Analysis of the blockchain reveals that the total number of accounts generating new blocks in NXT is between 300 and 350. Which is better than 100 delegates in BitShares.

[xeroc]

punishment == discouragement?
   

[bitbro]

BM said double spending only is a problem when it can be done anonymously (https://soundcloud.com/beyond-bitcoin-hangouts/hangout-6-28-2014-dan-larimer - 12:34) --> delegates as well as for mining pool operators are not anonymous(as least not in the sense that is relevant here) -> my assumption (since the OP) is that the only way to double spend is to mine / forge blocks in secret (assumption: miner/forger has more than 50%) and publish the then longer alternative chain to everybody later. While mining/forging in disguise the mining pool as well as the delegate can not collect the block reward --> miners will run away from the pool and the delegate will loose his votes --> reason why reversing the tx ledger with respect to a tx that lies back a while is practically not possible with POW / POS in so far the attacker doesnt have more than 50% himself as opposed to the assumed delegated voting power (mining pool operators and delegates).

Does this reasoning have any flaws? I think it is not difficult to answer. Just need someone else to think through it...

I would like to add that double spending is still a problem if it is EASY because then the cost of attempting many times can be low and the potential of getting away with many small anonymous transactions may be higher.   However, once you raise the threshold of attack to something that only someone who is already wealthy with a public reputation can pull off, then it is no longer a real concern because those that could do it have more to lose and are certain of getting caught.

When attempting to enforce "laws" all that matters is certainty of getting caught, not the severity of the punishment.

Last sentence does not seem to make sense BM


Sent from my iPhone using Tapatalk

[santaclause102]

Currently group owning 51% of the stake controls all 101 delegates.
They can:
Ignore transactions. Including banning single address from sending/receiving.
Rise transaction fees at will.
Have free transactions.

This is correct if the party really owns 51% of the stake. But if 51% of the delegates are evil and just got those 51% because they campaigned successfully for it then those delegates will loose their 51% again.

I would like to add that

that means you agree with my reasoning? It is really very basic. There is just no one explicitly agreeing or disagreeing since the OP... Anyone that agrees or disagrees with this https://bitsharestalk.org/index.php?topic=5330.msg73668#msg73668 ?

[emski]

Currently group owning 51% of the stake controls all 101 delegates.
They can:
Ignore transactions. Including banning single address from sending/receiving.
Rise transaction fees at will.
Have free transactions.

[bytemaster]

BM said double spending only is a problem when it can be done anonymously (https://soundcloud.com/beyond-bitcoin-hangouts/hangout-6-28-2014-dan-larimer - 12:34) --> delegates as well as for mining pool operators are not anonymous(as least not in the sense that is relevant here) -> my assumption (since the OP) is that the only way to double spend is to mine / forge blocks in secret (assumption: miner/forger has more than 50%) and publish the then longer alternative chain to everybody later. While mining/forging in disguise the mining pool as well as the delegate can not collect the block reward --> miners will run away from the pool and the delegate will loose his votes --> reason why reversing the tx ledger with respect to a tx that lies back a while is practically not possible with POW / POS in so far the attacker doesnt have more than 50% himself as opposed to the assumed delegated voting power (mining pool operators and delegates).

Does this reasoning have any flaws? I think it is not difficult to answer. Just need someone else to think through it...

I would like to add that double spending is still a problem if it is EASY because then the cost of attempting many times can be low and the potential of getting away with many small anonymous transactions may be higher.   However, once you raise the threshold of attack to something that only someone who is already wealthy with a public reputation can pull off, then it is no longer a real concern because those that could do it have more to lose and are certain of getting caught.

When attempting to enforce "laws" all that matters is certainty of getting caught, not the severity of the punishment.   

[santaclause102]

BM said double spending only is a problem when it can be done anonymously (https://soundcloud.com/beyond-bitcoin-hangouts/hangout-6-28-2014-dan-larimer - 12:34) --> delegates as well as for mining pool operators are not anonymous(as least not in the sense that is relevant here) -> my assumption (since the OP) is that the only way to double spend is to mine / forge blocks in secret (assumption: miner/forger has more than 50%) and publish the then longer alternative chain to everybody later. While mining/forging in disguise the mining pool as well as the delegate can not collect the block reward --> miners will run away from the pool and the delegate will loose his votes --> reason why reversing the tx ledger with respect to a tx that lies back a while is practically not possible with POW / POS in so far the attacker doesnt have more than 50% himself as opposed to the assumed delegated voting power (mining pool operators and delegates).

Does this reasoning have any flaws? I think it is not difficult to answer. Just need someone else to think through it...

[merockstar]

End The Fed Organizer in Germany Gets Car Fire Bombed:
https://www.youtube.com/watch?v=DucAAAA9yUw

I have seen that and I'm not disagreeing with you that "dissidents" to the status quo wouldn't be targeted; however, it's my opinion that delegate power is less influential than outward political opposition.  Delegates do have power but their power can *easily* stripped...in 24hrs or less if I'm not mistaken.  This means that targeting a delegate holds little value because another delegate would immediately replace them.

Ultimately, delegates equate to trust.  That trust could even be placed in politically neutral environments if need be.. or in international waters.  I wouldn't mind delegating from the open seas like Pirate Radio (their demise being that they weren't in international waters):
8)

idk i think this could be a legitimate concern. Suppose even just half of the delegates have actions taken against them. It would make it so that people would be scared to run a node. Only the most balls-havinest would step up.

[yellowecho]

End The Fed Organizer in Germany Gets Car Fire Bombed:
https://www.youtube.com/watch?v=DucAAAA9yUw

I have seen that and I'm not disagreeing with you that "dissidents" to the status quo wouldn't be targeted; however, it's my opinion that delegate power is less influential than outward political opposition.  Delegates do have power but their power can *easily* stripped...in 24hrs or less if I'm not mistaken.  This means that targeting a delegate holds little value because another delegate would immediately replace them.

Ultimately, delegates equate to trust.  That trust could even be placed in politically neutral environments if need be.. or in international waters.  I wouldn't mind delegating from the open seas like Pirate Radio (their demise being that they weren't in international waters):
8)

[onceuponatime]

(or assasination  :) )

End The Fed Organizer in Germany Gets Car Fire Bombed:
https://www.youtube.com/watch?v=DucAAAA9yUw

[emski]

worst case scenario, couldn't an anonymous pseudonym gain people's trust?

It might be needed as in a lot of countries you might not be permitted to run a delegate (or profit from one).

[santaclause102]

Can anybody answer the questions below?

Now is that

...the reason that this [a double spend] practically isn't possible with DPOS is that "When 51% of the delegates mine a hidden chain they can't mine the currently longest chain. This would [go public and] lead to them being 'unvoted'"?

true? And is there any difference in this respect compared to POW? I guess that all miner would leave a pool with 51% or more if the pool operator would begin to mine a hidden chain because then the pool couldnt mine the currenly longest chain anymore and make a profit. Correct?

[merockstar]

worst case scenario, couldn't an anonymous pseudonym gain people's trust?

[yellowecho]

(or assasination  :) )