Post
Topic
Board Speculation
Re: sidechains discussion
by
rocks
on 06/01/2015, 20:32:08 UTC
But unless the cartel has 100% of has power, which is impossible, these transactions would simply wait for an honest miner's block.

This is incorrect. A cartel with a high percentage could, within the protocol, reject blocks that don't follow its transaction inclusion rules. It doesn't take 100% to completely block some or all transactions from the chain.
A miner does not reject blocks, a miner can only build on top of a chain of blocks.

What you are describing is an attacking miner only building on top of their blocks and not on top of any blocks. That is the very definition of a 51% miner attack. The reason an attacking miner needs 51% is so their chain, that only includes their blocks with their inclusion rules, grows faster than the rest of the network.

There are two possible paths here.
1) There are two competing chains where one chain has only the attacker's blocks and the other chain has the honest miner's blocks. The goal of the attacker is to make their chain with their inclusion rules the longest. The network here can optionally choose the honest chain (which is what Gavin's proposal does). It also require 51% to pull off, which I don't think is possible anyway.
2) The blocks from both the attacker and honest miners build on top of each other in a single chain. In this this situation the honest miners include all transactions skipped by the attacker, and the attack fails.