Implement Gavin Andresen's defence

[IanMichaelHarper]

https://gavintech.blogspot.com/2012/05/neutralizing-51-attack.html

[IanMichaelHarper]

https://github.com/mikekelly/bitcoin/blob/22261599c3ae0f55421a4bac66371fd13c6036a1/src/validation.cpp#L1412

Might be useful for implementation.

[mikekelly]

I'm not sure that @gavinandresen proposal works. How can it be consensus compatible if it depends on nodes knowing the order in which blocks were announced? Wouldn't it cause new nodes to be immediately chainsplit (since they will pick the heaviest chain)?

[IanMichaelHarper]

How can it be consensus compatible if it depends on nodes knowing the order in which blocks were announced?

Well PoW is still used in the minority chain to order blocks, but it seems this change looks at competing chains rather than competing blocks. Nodes would probably measure the sums across multiple blocks before orphaning the heaviest chain.

Wouldn't it cause new nodes to be immediately chainsplit (since they will pick the heaviest chain)?

I imagine new nodes would download the latest software that would include this new rule. If they are using old software then yes they would remain on the heaviest chain. I'm pretty sure this rule would create a hard fork.

One thing I'm not 100% about is that when @gavinandresen says "priorities of transactions". I assume he means tx fees, but if that's the case idk why he doesn't just say that.