4 ms·
Now the million dollar question: Does price follow hashrate, or hashrate follow price?
by kloch 4y ago
Now the million dollar question: Does price follow hashrate, or hashrate follow price?
- alchemist1e9 4y agohttps://coinmarketcap.com/currencies/ethereum-pow/ https://coinmarketcap.com/currencies/ethereum-pow/ This IOU market seems to be suggesting there will be a new PoW chain that has higher market value than ETC. That implies the miners are expected to split their current hashing power between ETC and this new one. However I have not found the necessary organizational effort to create this new PoW chain, so with hours to go, it’s a bit of a mystery to me, as to why this IOU market is trading as high of a price as it is.
- TakeBlaster16 4y agoYour comment made me expect it to be a lot higher than it was. I remember the BCH fork trading at 10-20% of BTC for months and peaking as high as 50%. 2% seems like a more appropriate ratio.
- alchemist1e9 4y agoThat’s definitely true. However technically it would have a larger total market cap than ETC at current prices.
- SilasX 4y agoYep, this is why I moved my ETH off exchanges for the merge, so I could recover the PoW ETH (ETHW) in case the exchange won't support that branch. However, there's also the problem that if I broadcast an exchange on ETHW, someone can replay it on the PoS chain (ETHS), meaning, if I wanted to sell both branches on an exchange, I'd have to find one that let me send both to the same address. Unless I'm horribly misunderstanding something.
- syntheticcdo 4y agoIf you want to touch the POW chain, you should move all of the money out of the identical wallet on the POS chain first to avoid a replay attack. If the POS wallet address is empty (post split) there is no chance of replay because the address is already empty. Of course, just as possible is a replay of the POS transaction on the POW chain, so the POS wallet chosen for the initial emptying should be 100% controlled by you as well (ie- not an exchange wallet).
- SilasX 4y ago>Of course, just as possible is a replay of the POS transaction on the POW chain, so the POS wallet chosen for the initial emptying should be 100% controlled by you as well (ie- not an exchange wallet). Huh? That's the exact scenario I described, and it doesn't avoid the problem: say I empty the PoS wallet onto an exchange post-split. Then a joker can replay it on PoW, emptying my PoW wallet. The ETH on PoW goes to the corresponding PoW address, which I may or may not be able to benefit from. I don't know what problem you solved here from the original comment.
- syntheticcdo 4y agoHere's the full breakdown: 1. You have 100 ETH in original ETH chain in wallet A. 2. Merge occurs, and a forked POW chain appears 3. You now have 100 ETH in wallet A, and 100 ETH-POW in wallet A. 4. Generate a new wallet B (note that because the protocol for ETH and ETH-POW is identical, wallet B is valid for both ETH and ETH-POW). 5. Transfer 100 ETH from wallet A to wallet B (on POS chain). 6. Observe both wallets on both chains: Scenario A: There is no replay in ETH-POW: - You have 100 ETH in Wallet B - You have 100 ETH-POW in Wallet A Scenario B: Your ETH transaction is replayed on ETH-POW - You have 100 ETH in Wallet B - You have 100 ETH-POW in Wallet B In scenario A, you are free to do whatever you want with the 100 ETH-POW in wallet A -- either move it to an exchange that supports the ETH-POW chain, or continue using the ETH-POW chain. There is no risk of replay because wallet A on the POS chain is already empty. In scenario B, you still have both 100 ETH and 100 ETH-POW in wallet B which you control. Write off the 100 ETH-POW because the risk of interacting with it is likely much greater than any value the coins in that chain have. In both scenarios, your 100 ETH is safe and controlled by you.