POW is just POS but where instead of token holders deciding the fate of the chain it is done by miners with the biggest fiat wallets that can purshase hardware. And then you have the whole scalability and high fee issues that POW has. There is a reason why Ethereum is on the path to transfer to POS.
You can't avoid politics, you either design the chain with on-chain governance or you leave it with one that is ad hoc that can only be resolved via contentious forks, hashing wars and 51% attacks.
The way I see it the main issue with the current DPOS implementation is that the only way to protect the chain is with long staking periods that gives little incentive for money attacks. The whole Stinc debacle could have been avoided if the system was designed to avoid minority attacks.
The 30 day governance waiting period that was included in the last hive HF is a stop gap. You "only" need 65.5 million hive to take over consensus which is about half of what is on the exchanges. The main design changes that I would make to DPOS is to remove the 30 vote rule and change it to a "one hive = one vote" system and give an economic incentive to participate in governance by only paying staking rewards to the accounts that vote for block producers.
In such a system you could still vote for several witnesses but each additional vote would decrease the strength of each one instead of multiplying it by 30 like we have today.
You're thinking of public chain settings using standard release builds where anyone can come in and do POW hashing provided they have your network ID. The HSC isn't going to provide that ability one in part because the EVM executable itself is going to be a heavily modified instance that won't even be able to understand classic ETH clients and also because there are ways to mitigate the advantage that having a large amount of hashpower will yield miners depending on how you configure the chain to select block miner. It's doable to have a properly decentralized private EVM chain setup between PoW users without favouring hashrate and by going this route by the look of it we should be able to avoid a lot of the classical failures of small PoW networks.
You can avoid politics if you make it trustless and human interaction completely optional. Because HSC v1.0 will actually serve as a means of burning HIVE rather than adding to inflation I think you'll see that it ends up actually attracting more decentralization as mining clients become available allowing even low power devices the chance to hash blocks to earn some HIVE which comes from the gas fees paid in HIVE.
I'm not making anything DPOS on this sidechain.. There isn't going to be a risk of 51% attacks or lots of the other classical ETH downfalls due to the EVM binary itself changing to a point where it can't be directly compared to existing popular EVM capable chains. It's going to be of based around an already existing chain with it's own set of rules, It merely adds more ways to utilize your HIVE, but not a free transaction like base layer chain operations.. You pay the sidechain to pay the miners to mine the blocks to do your decentralized function call or whatever. It all runs on HIVE, the total HIVE token cap is slightly reduced by burning some percentage of the gas fee to @null. Most goes to the miner of that block then a sustainable amount goes to a developer fund for future R&D.
DPOS used to be qutie strong in my eyes but Sun Yuchen totally spooked me off it being bulletproof. :/
Yeah. In the event that HIVE gets taken over again we just fork and do our own thing again.. :(
Now you peaked my interest, it will be very interesting to see how the game theory plays out which such a design. Thanks for the detailed explanation.
No problem! You certainly raised some very valid points though for anyone just going in and cloning ETH without much thought to how it needs to operate given it's determined goal. Much appreciated!