>there have been no foundational attacks against important PoS systems in the wild, whereas there have been multiple successful attacks against medium-size PoW systems.
Considering that the largest PoS blockchains are currently securing billions of dollars of value, why has no adversarial attack occurred? Surely this proves that PoS is a secure consensus mechanism to rival PoS at a fraction of the cost?
Corpo coins that are willing to fork at the first sign of attack haven't had any attacks? They must be just as good. Just wait until Governments realize the power they hold over such networks - that's where this simplistic argument falls apart.
Lucas Martin
What can government do to PoS that it can't do to PoW?
Ryder Cook
>sanctions all of your wallets for using a homosexual racist anonymous tranny forum
Nathan Cox
>PoS is a secure consensus mechanism Proof of Work and Proof of Stake are Sybil Control mechanisms, by itself they say nothing about security, decentralization or efficiency. Its basically just Spam Protection.
>exchanges control 60% >government forces exchanges to implement changes How is that not an external attack?
Brody Jenkins
Because the idea that you can buy infinite eth without the price moving or without anyone noticing is as much of a meme as the idea of a new PoW chain
Jeremiah Gray
Problem with PoS, at least in Ethereum's case, is that most validators are/will be exchanges or companies (Lido, Binance, Kraken, etc.) and as such they are liable to regulatory pressure from OFAC and other regulatory bodies. In theory they could very well censor some transactions.
Chase Young
>bunch of shitty coins fork bitcoin and copy exact same PoW consensus algorithm >exact same asics for bitcoin mining also work for shitty forked coin > miners occasionally turn their rigs on forked coin and easily have more hash power
This doesn't mean all PoW coins are vulnerable, only the ones which copy the exact same consensus algorithm as much bigger PoW chains which means there are already tonnes of Asics already manufactured and in use which can be used on the much smaller coin.
any PoW coin even if it only had 1 millionth of the hashing power securing it as bitcoin wouldn't be that vulnerable to 51% attacks if it required different asics to mine from bitcoin or other much bigger PoW coins.
typical pozzed PoS disinfo.
Cameron Torres
other than the mask
Owen Russell
no. bitcoin is inefficient, PoW is not. a PoW network that can actually scale further than bitcoin's 7tps turns the debate into a nonissue. then you get PoW security without PoS pozzed complexity and censorship issues.
Sebastian Nelson
Whales buy OTC lol, is this your first day? Why is Bitcoin inefficient? Censorship protection on PoS chains is abysmal. 2/3 of validators are enough to collude and censor transactions meanwhile in PoW a small amount of miners is sufficient to include your transaction. See my comment above. See this as well: reddit.com/r/dot/comments/qdj64e/we_dont_have_a_full_control_of_our_tokens/
Jose Campbell
STFU SCHIZOS AND BUY THE NEXT MARK CUBAN ANIMAL SCAM
Nathan Turner
huh? consuming electricity the size of norway's grid for just 7tps? this is not a global payment system, it's just beginner technology. but if you scale PoW properly for higher throughput, the energy consumption becomes a moot point because energy/tx is worth it compared to the current systems. satoshi laid out something good, but it has to evolve and innovate
(You) can, the world can't. I'm all for alternate forms of hashing, that's also included in the innovation angle. It has to run on renewables anyway in the future. The point is that PoW can actually be scaled higher than 7tps, on the base layer.
Lightning and other L2's are a meme and don't count as actual scaling. It has to be base layer, which is possible by the way.
James Collins
Sure, it's possible in the short or mid term but then you forsake full nodes. What's going to take to run a full node if Bitcoin were to increase block size significantly? Take a look the the node requirements for BitcoinSV lol bitcoinsv.io/svnode
64gb RAM? Who on earth has 64gb of RAM on their laptop? This is why increasing the block size is a bad idea.
Christian Torres
>Censorship protection on PoS chains is abysmal. 2/3 of validators are enough to collude and censor transactions meanwhile in PoW a small amount of miners is sufficient to include your transaction. In pow a majority is enough