You are viewing a single comment's thread from:

RE: Democratizing Steem!

in #steem5 years ago

http://vukolic.com/iNetSec_2015.pdf For example talks about the issues of mining pools leading to centralization. Which got it mainly from here: https://www.infoq.com/articles/is-bitcoin-a-decentralized-currency/ which describe most mining pools as a threat and hope that more decentralized mining pools surge (Where you can apply the mining pool to big stakeholder issue 1:1 almost).
I can't find the paper atm which was talking about the monopoly problem.
http://www.coinfox.info/news/reviews/6417-proof-of-work-vs-proof-of-stake-merits-and-disadvantages
Does summarize it pretty well though.

If there is a single stakeholder which MAY define all witnesses and then almost do whatever he wants with the chain. Then this is a major risk for the security of the chain. And thus it is not very well decentralized. I don't know what it so difficult to understand with this.

Imagine, someone from another Blockchain doing this to bring down Steem to value his own Blockchain or Steem fork. At the moment this would be even "relatively cheap".

Sort:  

First link references

https://cacm.acm.org/magazines/2018/7/229033-majority-is-not-enough/fulltext
For their assertion that bitcoin mining pools can undermine the network, but I'm not going to spend one cent let alone 8 thousand to see exactly what the "greedy" miners can do to undermine the network, so I'll cautiously point out that the 51% attacks have been possible for years and we still have no proof of any of that happening on bitcoin. Even if it's a possibility, it's in the realm of counter to the incentive of mining/securing the network, and if it does happen it likely will be reversed, new rules put in place and things will ge back to working in no time.

As for the other article, it's complete horseshit. As Vitalik said, has there been any such attacks on POS.
No

And that's 3 years later and still nothing to show for it.

Let's suppose for a moment that there's ANY merit to this:

If there is a single stakeholder which MAY define all witnesses and then almost do whatever he wants with the chain. Then this is a major risk for the security of the chain. And thus it is not very well decentralized. I don't know what it so difficult to understand with this.

Imagine, someone from another Blockchain doing this to bring down Steem to value his own Blockchain or Steem fork. At the moment this would be even "relatively cheap".

So theoretically, because of the size of someone's stake, they "can" do whatever they want with the chain, saying essentially that with their stake alone, they could force everyone else to comply. Well, good luck getting them to undermine their position so that your unfounded fears and concerns will be assuaged, because in reality, not only is it a complete fantasy to think that one stakeholder could do such things, but it undermines the entire reason for holding stake to penalize, to tax, and to otherwise undermine the power of the stake in relation to how much it is, thereby making holding a certain ammount of stake a risk without any reward, and exactly like a pool of miners colluding to cheat a few bitcoin while essentially destroying the value of bitcoin it's the same for one large stakeholder to invest into the system only to undermine it, because sinking one ship in your head will raise the level of the ocean.

Following that logic, Facebook is not centralized either. There is one shareholder who has mayor stake in the company and which pretty much acts in "the best interest" of the company because of that. So, why do we need cryptocurrencies. Can't we just trust Paypal and Facebook? They are acting in the best interest of their companies too. They're kind decentralized as well. Got a lot of people from all over the world holding their stocks. Let's let Paypal put their money on the blockchain and let them handle the permissioned ledger and it's going to be alright.

The reason I'm in blockchain is exactly to not have this situation. On Steem there was recently a situation where some Witnesses thought about forking the chain, and Steemit Inc said they could elect just the 20 witnesses themselves in that moment.

There is the possibility of that happening, and they said they would do that. Is that decentralization or does it just kinda seem like it.

And of course, as long as they got 30 votes to cast, they could just do that, with 10 votes, especially on the short term, that is much more difficult.

I'm not saying that huge stake holders shouldn't be able to have a big influence. And I'm not denying that this makes sense in the context of game theory. Nonetheless, on Steem this influence is essentially to big, which is a big risk for the chain. So reducing this to 10 votes, would improve the situation while still maintaining the central elements of PoS. (And no, I don't want 1 vote, I want 10).

It's not the same thing by any stretch of the imagination:

No infrastructure is compromised by any stakeholder, and no stakeholder could compromise it by playing the game of voting for witnesses, which is largely a popularity contest and has absolutely no basis in reality as a point of failure, and what they would need for that to happen is for witness to go along with them, which again is countered by common sense, as compromising the network will not gain them anything, and even if the large stakeholder compromises the network by voting their own self as top witnesses they would still have to get the rest of the stakeholders to go along with it, and that's not feasible, as by that point they would have went against the witnesses and the majority of large stakeholders if not the entirety.

The whole thing is bogus: there's absolutely no problem, as that article points out when it removed both pumpkin and freedom's stake in witnesses standing, which I'm sure was never their intention, they probably wanted to fan the fears that you're trying to as well: toooooooo much stake.

I'm not saying that huge stake holders shouldn't be able to have a big influence. And I'm not denying that this makes sense in the context of game theory. Nonetheless, on Steem this influence is essentially to big, which is a big risk for the chain. So reducing this to 10 votes, would improve the situation while still maintaining the central elements of PoS.

Yeah, too big, the point is that no such thing exist (risk) but even if it would, it basically says to people "you can invest only this much or you're nerfed", which is essentially counter to the very notion of holding stake for almost everyone: to accumulate ever more influence, and ultimately you're trying to make changes that are completely counter to the large stakeholders, and if you don't trust them to safeguard their investment which incidentally means to not compromise the network it doesn't make any sense to trust a buch of people who have monumentally less stake than them to do the same. Then there's the obvious implications of dissuading stake concentration, and making abuse harder to track since now there's 3 accounts instead of 1, or 30, which is why the native option to vote makes sense at 30, as that is the number of influential witnesses responsible for adapting new changes, and trying to mitigate people's influence over it is essentially square dab in the middle of the road to hell (stake splitting) being paved with good intentions.