1 min readDec 11, 2018
- Yes it could be use select between chains in the presence of a 51% attack. The issue there is POW is already overwhelmed by the attacker so you couldn’t use POW for the anti-sybil mechanism. You’d need to use utxos or coinage instead.
- Full nodes could technically also query the consensus group to see which txs are invalid though I don’t think it would be necessary (though it would be if POW is not used for anti-sybil). There could theoretically be a higher instance of orphans but in practice wouldn’t expect to see that as miners would not want to lose blocks by not using avalanche.
- Yeah that’s right. With the current pool structure, very few actual people would be part of the consensus group. Pool centralization is something I want to address at some point. Imo I think we can get it to where individual miners all use their own nodes and pools only aggregate and distribute the reward.
- So I think if a block contains a never-before-seen double spend you need to still do avalanche between the conflicting transactions. If avalanche says the tx in the bock is invalid then the block is invalid. If you think about it, there’s very little difference between receiving a block with a never-before-seen double spend in it and receiving an unconfirmed double spend just milliseconds before receiving a block containing it. In the later case you would obviously need to still do avalanche. So probably in the former case as well.
