Profits, Performance, Trust & Efficiency

[Agent86]

The trustee has only one power, to produce blocks that conform to the rules (or not to produce said blocks).
...
Trustee cannot change the blockchain rules.
Nodes do not trust the Trustee but validate every single transaction.
Trustee cannot create alternative chains or get away with double spend attacks.

This seems to make sense to me.

On some level it seems that a successful DAC will still need people who take on leadership roles (and could be bad if we don't have such people).  Every crypto has centralized development.  Being able to delegate your voting rights / stake to someone you trust who in turn can delegate further may be as good a way as any.  Maybe if Keyhotee is more developed it also becomes easier to build trust for individuals?

[clout]

I don't know if its just me but I feel like so much of this lies in theory and not in evidence. Why wouldn't you try to see how tapos + mining works, before you discard it.

[toast]

Did you see my delegates with revocation certificates suggestion? Comments?

Effectively 'delegation' using your stake is exactly what I am proposing.   Person A delegates to person B, who in turn can delegate to person C.   Eventually all of the delegates have to arrive at a 51% consensus on who produces the next block.   If someone sets up some software in a 'set it and forget it' manner then the network is perfectly safe unless this person is attacked.   If they are attacked then the network can quickly appoint (probably pre-arranged) the fallback. 

This is exactly what I'm interested in, how can we make this failover more automatic or not really look like failover (where I was going with nxt style queue).
I want to give a preference ordering on trustees, weighted by stake I'm willing to give.

[bytemaster]

Did you see my delegates with revocation certificates suggestion? Comments?

Effectively 'delegation' using your stake is exactly what I am proposing.   Person A delegates to person B, who in turn can delegate to person C.   Eventually all of the delegates have to arrive at a 51% consensus on who produces the next block.   If someone sets up some software in a 'set it and forget it' manner then the network is perfectly safe unless this person is attacked.   If they are attacked then the network can quickly appoint (probably pre-arranged) the fallback. 

At any time someone with stake can change who they delegate their vote to. 

If we can stay objective here and identify what attack we are afraid of and define what we are attempting to defend against in the most clear way possible then we can find the best solution.   Get to the *root* of the problem and solve that.

Without getting to the root we have solutions in search of problems and we end up paying for things that we do not need or even want (mining). 

[bytemaster]

I don't know that I completely understand the power that the trustee(s) have and what happens if all trustees colluded.  But, I'm willing to reserve judgement if you think it's the best system and give it a shot.

I think that if the mechanisms are not well communicated there would be a lot of FUD that would affect initial demand/value.

The trustee has only one power, to produce blocks that conform to the rules (or not to produce said blocks).
The trustee's power is delegated from the shareholders who can revoke it at any time and appoint a new trustee.
Redundancy is possible in case trustee is compromised.

Trustee cannot change the blockchain rules.
Nodes do not trust the Trustee but validate every single transaction.
Trustee cannot create alternative chains or get away with double spend attacks.

Without a trustee all other technical solutions are attempts to do the following:
1) Randomly select someone proportional to their stake in the system... they become 'trustee for a block'.
2) Pay / Punish this person for not producing a block.

As soon as you have this random selection you simultaneously open up the network for random attacks by anonymous parties.  The sheer number of hypothetical attacks that are possible under these 'random selection' processes means that the network is likely less secure and predictable.   All of a sudden people start mining blocks that exclude bids/asks to their benefit or to trigger margin calls.   Put another way, randomly selecting people to produce blocks gets you 'average' performance and 'random' manipulation.   Delegating through your proof of stake to a trustee maximizes the predictability and fairness of the chain. 

 

Remember, true decentralization is competition and reduction of barriers to entry.   

[toast]

With proof of stake your private key is your proof.   


Sent from my iPhone using Tapatalk

Do you mean private keys have to be risked with POS (forging) because they can not be used for forging when they are in cold storage?

They don't if you can sign for a delegate! Keep 90% in cold storage and have your 10% hot wallet sign for all of them. You could easily have revocation keys so you wouldn't have to take your shares out of cold storage to take away the voting power if your hot wallet is swiped.

I was thinking something similar.  Why couldn't you link your address to another address that has zero coins assigned to it.  You could then put your real wallet into cold storage and the empty wallet could provide security through POS.  If someone hacks my PC they would only get my voting power, and only until I pulled my wallet out of cold storage and used it to sign a new block with a new link.  Am I making sense?  Could this work?

You don't need zerocoin for this, but the idea is similar to a zerocoin "ticket". You just put a hash of a serial in the same transaction where you sign the delegate, and the rule is that if the network sees the original serial then it considers the delegation revoked. This way you can take away voting power without taking your shares out of cold storage. Voting for a new delegate signer would require you to move the cold shares though.

If there is an established means of providing security through PoS that doesn't involve having to leave your wallet online constantly, why not take that approach?

Did you see my delegates with revocation certificates suggestion? Comments?

[toast]

I don't know that I completely understand the power that the trustee(s) have and what happens if all trustees colluded.  But, I'm willing to reserve judgement if you think it's the best system and give it a shot.

I think that if the mechanisms are not well communicated there would be a lot of FUD that would affect initial demand/value.

A trustee is just a node that asserts it saw a block by signing it. It stops forks, but gives the power to stall the network OR perform a double-spend with cooperation of a large number of shares. They would get caught though, which is actually really important.

[puppies]

With proof of stake your private key is your proof.   


Sent from my iPhone using Tapatalk

Do you mean private keys have to be risked with POS (forging) because they can not be used for forging when they are in cold storage?

They don't if you can sign for a delegate! Keep 90% in cold storage and have your 10% hot wallet sign for all of them. You could easily have revocation keys so you wouldn't have to take your shares out of cold storage to take away the voting power if your hot wallet is swiped.

I was thinking something similar.  Why couldn't you link your address to another address that has zero coins assigned to it.  You could then put your real wallet into cold storage and the empty wallet could provide security through POS.  If someone hacks my PC they would only get my voting power, and only until I pulled my wallet out of cold storage and used it to sign a new block with a new link.  Am I making sense?  Could this work?

You don't need zerocoin for this, but the idea is similar to a zerocoin "ticket". You just put a hash of a serial in the same transaction where you sign the delegate, and the rule is that if the network sees the original serial then it considers the delegation revoked. This way you can take away voting power without taking your shares out of cold storage. Voting for a new delegate signer would require you to move the cold shares though.

If there is an established means of providing security through PoS that doesn't involve having to leave your wallet online constantly, why not take that approach?

[Agent86]

I don't know that I completely understand the power that the trustee(s) have and what happens if all trustees colluded.  But, I'm willing to reserve judgement if you think it's the best system and give it a shot.

I think that if the mechanisms are not well communicated there would be a lot of FUD that would affect initial demand/value.

[toast]

With proof of stake your private key is your proof.   


Sent from my iPhone using Tapatalk

Do you mean private keys have to be risked with POS (forging) because they can not be used for forging when they are in cold storage?

They don't if you can sign for a delegate! Keep 90% in cold storage and have your 10% hot wallet sign for all of them. You could easily have revocation keys so you wouldn't have to take your shares out of cold storage to take away the voting power if your hot wallet is swiped.

I was thinking something similar.  Why couldn't you link your address to another address that has zero coins assigned to it.  You could then put your real wallet into cold storage and the empty wallet could provide security through POS.  If someone hacks my PC they would only get my voting power, and only until I pulled my wallet out of cold storage and used it to sign a new block with a new link.  Am I making sense?  Could this work?

You don't need zerocoin for this, but the idea is similar to a zerocoin "ticket". You just put a hash of a serial in the same transaction where you sign the delegate, and the rule is that if the network sees the original serial then it considers the delegation revoked. This way you can take away voting power without taking your shares out of cold storage. Voting for a new delegate signer would require you to move the cold shares though.

[phoenix]

I feel like with TaPOS you need a lot of people to keep their keys online to secure the network, but then they might be hacked and lose their shares. With a trustee, you need enough people to be interested in voting to achieve that 51% threshold, and you need a way to determine who everyone is voting for that risks as little as possible.

[puppies]

With proof of stake your private key is your proof.   


Sent from my iPhone using Tapatalk

Do you mean private keys have to be risked with POS (forging) because they can not be used for forging when they are in cold storage?

They don't if you can sign for a delegate! Keep 90% in cold storage and have your 10% hot wallet sign for all of them. You could easily have revocation keys so you wouldn't have to take your shares out of cold storage to take away the voting power if your hot wallet is swiped.

I was thinking something similar.  Why couldn't you link your address to another address that has zero coins assigned to it.  You could then put your real wallet into cold storage and the empty wallet could provide security through POS.  If someone hacks my PC they would only get my voting power, and only until I pulled my wallet out of cold storage and used it to sign a new block with a new link.  Am I making sense?  Could this work?

[toast]

With proof of stake your private key is your proof.   


Sent from my iPhone using Tapatalk

Do you mean private keys have to be risked with POS (forging) because they can not be used for forging when they are in cold storage?

They don't if you can sign for a delegate! Keep 90% in cold storage and have your 10% hot wallet sign for all of them. You could easily have revocation keys so you wouldn't have to take your shares out of cold storage to take away the voting power if your hot wallet is swiped.

[wasthatawolf]

2) Mining results in a kind of centralization and de-facto trustee that cannot be fired

This is misleading.  With this comparison, there would only be a singular "de-facto trustee" if someone controls 51% of the network.  Otherwise the "de-facto trustee" is the collective group of miners.  If you're referring to the pool operators as the "de-facto trustee" that assumes that they control the net mining power of their pool.  We've already seen with bitcoin that when pools get too big, people will stop mining within them.

With a trustee, you're talking about centralizing control, hiding it behind a Tor node, and leaving it up to the shareholders to fire the trustee if they think they're acting to harm the network.  How would you even elect a trustee in the first place?  Campaigning?

[santaclause102]

With proof of stake your private key is your proof.   


Sent from my iPhone using Tapatalk

Do you mean private keys have to be risked with POS (forging) because they can not be used for forging when they are in cold storage?