Asics is able to destroy Grin network and go to MWC it was specially created by ASIC Maker for this reason. Asic is not about saving - the reason is they about centralization not decentralization.
Monero and ETH not have problem 51% and asic problem. Asic it is not processor. Asic it it destroy of Satoshi consensus 1voice - 1 processor.
And those that not understand this truth is fulls.
Most powerful vector 51% attack is maining pools.
If we put anti pool rules we will secure the network .
Anti pool rules is simple. For example: Right to finalize block 1node - 1time - per 1day. It provides decentralization. But now some few nodes have together all grin blocks.
Do you even know which node was the original block propagator in a p2p network? Even if you did, a node is not 1 person. One person can run 1000 nodes.
People will be interested to have more then 1 node the network will be stronger and decentralized it is good. Some one with 1000nodes is nothing against crowd of millions people with 10-100nodes each.
Rules for nodes is simple to organaze in code logic.
With the latest Ethereum Classic attacks of 51% I was thinking a little about a possible solution.
Voluntary Guarding with Proof-of-Stake
Hybrid Proof-of-Stake/Proof-of-Work coins like Dash and Decred are mostly regular Proof-of-Work coins. In addition those coins allow coin holders to lock their coins for some time to participate in a voting process. Votes are just specialized transactions. Coin holders are motivated to participate in voting by a part of block reward. Miners are motivated to include votes in blocks. It is enforced on the consensus level.
I believe such approach can be achieved in purely voluntary manner with virtually no consensus changes. Coin holders can register “guards” by revealing amounts with a fixed denomination (say 1000 GRIN), then based on block hash lottery they will be selected for producing voting transactions (few guards for a block). Better if the process relies on timelocks too. Miners can include these voting transactions with a zero fee. Nodes, including miners, can monitor guard participation on the chain and make decisions based on it: show a warning if drastic changes in guard participation occur, reject deep chain reorganization with low guard participation and so on.
Pros:
- Mining process won’t be affected at all as well as ASICs development. 100% of the block reward go to a miner;
- May be an effective protection against 51% attacks;
- Leads a way for better chain upgrades;
- Guards are motivated not by staking rewards but by positive feedback on price as providing network safety;
- Does not motivate to buy and lock coins for rewards and should not harm economical properties.
Cons:
- Revealing amounts can lower the overall privacy (but as it purely voluntary, participation level probably won’t be very high);
- May have specific issues on MimbleWimble because of cut-through;
- Requires additional transactions in a block;
- Motivation to run guards might be not sufficient.
Additionally:
- Better to find a way how to protect guard funds from stealing by “cold” voting or voting delegating.
- Voting process probably can be done off-chain as it is not a part of consensus while guard registrations rather to be done on-chain.