BitShares Forum
Main => General Discussion => Topic started by: xeroc on July 06, 2014, 06:53:21 pm
-
http://www.reddit.com/r/Bitcoin/comments/29yyjm/proofofstake_and_distributed_consensus_are/
Vitalik on reddit:
That's not quite what I said. I said that every pure PoS system, except a PoS system with trusted chain download sources, can always and forever be taken over by a 51% collusion of the genesis block allocation. If you're willing to accept that, then TaPoS-based systems plus slasher plus some kind of yet-undiscovered transaction aggregation protocol seem to be quite viable, albeit flimsier in certain ways.
Whats TaPoS in ethereum language?
-
Transactions as proof-of-stake... referencing recent block hashes in your transactions to prevent reversing history too far. Dan's brainchild ;)
-
Seems like others are having a hard time keeping up with our latest developments as his recent article made no mention of DPOS. That is perhaps a failure on our part.
-
Seems like others are having a hard time keeping up with our latest developments as his recent article made no mention of DPOS. That is perhaps a failure on our part.
V's response to DPOS:
[–]vbuterin 1 point 5 minutes ago
I would say its resistance to long-range NaS is the biggest concern, and that's dependent entirely on the implementation. Theoretically DPOS is something that could be layered onto any other PoS algorithm; it's actually a slightly higher-level construction than naive PoS, slasher, TaPoS, etc.
-
Seems like others are having a hard time keeping up with our latest developments as his recent article made no mention of DPOS. That is perhaps a failure on our part.
V's response to DPOS:
[–]vbuterin 1 point 5 minutes ago
I would say its resistance to long-range NaS is the biggest concern, and that's dependent entirely on the implementation. Theoretically DPOS is something that could be layered onto any other PoS algorithm; it's actually a slightly higher-level construction than naive PoS, slasher, TaPoS, etc.
What is long-range NaS?
-
nothing-at-stake letting you build a chain from a long time ago. Like he said, it depends on the implementation and how you provide the underlying long-term security. For us the "difficulty" metric is total chain approval. Again like he said, it is still vulnerable to collusion from 51% of initial stakeholders.
-
I made some posts if you can give me an up vote!
-
I made some posts if you can give me an up vote!
Sure
+5%
-
I made some posts if you can give me an up vote!
I gave you one; Looks like they could use a few more
http://www.reddit.com/r/Bitcoin/comments/29yyjm/proofofstake_and_distributed_consensus_are/
-
Seems like others are having a hard time keeping up with our latest developments as his recent article made no mention of DPOS. That is perhaps a failure on our part.
V's response to DPOS:
[–]vbuterin 1 point 5 minutes ago
I would say its resistance to long-range NaS is the biggest concern, and that's dependent entirely on the implementation. Theoretically DPOS is something that could be layered onto any other PoS algorithm; it's actually a slightly higher-level construction than naive PoS, slasher, TaPoS, etc.
What is long-range NaS?
Long-range NaS assumes 51% founders have nothing to lose from undermining the chain they created. It also assumes that the users of the system are robots with no ability to adapt. Real consensus exists outside the blockchain and the active delegate slate (elected post-genesis) would clearly implement automatic checkpointing anytime participation is high and therefore they wouldn't even look at alternative chains that date back to genesis.
All the more reason to have wide allocation in genesis like we have with AGS/PTS.
Bottom line, V. has just demonstrated the superiority of DPOS by showing the weaknesses in the other systems.
-
I made some posts if you can give me an up vote!
I gave you one; Looks like they could use a few more
http://www.reddit.com/r/Bitcoin/comments/29yyjm/proofofstake_and_distributed_consensus_are/
+5%
-
I made some posts if you can give me an up vote!
Sure
+5%
done
-
Long-range NaS assumes 51% founders have nothing to lose from undermining the chain they created. It also assumes that the users of the system are robots with no ability to adapt.
A lot of these anti-POS arguments are based on an assumption of exceptional irrational behavior. So yes in POS you can mine to 2 forks in parallel but why would anyone do this ? The NXT guys called him on this too. Vitalik references altruistic POS block-signers preventing this but it isn't "altruistic" block-signers, they're just block-signers/miners who don't want to destroy the network. That isn't altruism.
Basically there are 2 parties to me. Those for network security and those trying to undermine it. Those trying to undermine it will have to buy a stake and then destroy it while destroying the network. (In a general sense.)
With Vitalik there are "self-interested miners" who would try to undermine the network that they're invested in. The good guys are the altruistic miners. It doesn't really jive with my version of reality, but it is a great way for him to frame his argument. I mean.. who isn't "self-interested".
-
Basically there are 2 parties to me. Those for network security and those trying to undermine it. Those trying to undermine it will have to buy a stake and then destroy it while destroying the network. (In a general sense.)
There's incentive for the existing financial status quo, Central Banks & Banks who are threatened by the rise of crypto-currencies and crypto-equities, to undermine & destroy the networks & thus the competition, the cost isn't too relevant/significant to them at the early stages. No?
-
Seems like others are having a hard time keeping up with our latest developments as his recent article made no mention of DPOS. That is perhaps a failure on our part.
Marketing has been the soft underbelly of your company. Your products are so innovative that it's difficult for a critical mass of consumers to grok in a flash. Once BTSX is launched, marketing is going to become a whole lot easier.
BTW, Bytemaster, thank you for everything you do. I appreciate the depth of your vision, your transparency, your work ethic, and your ability to conjure a team of talented, like-minded developers over just a few months. Of all the bitcoin 2.0 projects, it's pretty obvious I picked the right one to invest in. Thank you.
-
Basically there are 2 parties to me. Those for network security and those trying to undermine it. Those trying to undermine it will have to buy a stake and then destroy it while destroying the network. (In a general sense.)
There's incentive for the existing financial status quo, Central Banks & Banks who are threatened by the rise of crypto-currencies and crypto-equities, to undermine & destroy the networks & thus the competition, the cost isn't too relevant/significant to them at the early stages. No?
I suspect that POS is still more expensive than buying 51% of network hashes. It would be interesting to see this as a study across all manner of altcoins - their network security cost vs their market cap. With POW you can't fork miners out as an option either.
Then as Dan points out it is the ultimate in free-market, so the POS network can be forked with the malicious entity left out. So everyone is back to square one and the attackers have lost their investment. Sure they'd spread FUD, but the network would continue on with significant equity concentration.
-
I suspect that POS is still more expensive than buying 51% of network hashes.
Now, when u can rent hashing power for a day, PoW coins r very insecure.