Chia has weaker guarantees than Nakamoto consensus.
Every proof-of-work solution in Bitcoin commits to the transaction history,
but the proofs-of-space in Chia only commit to the farmer public key.
Which is used to separately sign for a block of transactions.
Which allows a collusion of recent farmers to produce alternate
transaction histories at no cost.
Ah, thanks for the link! It seems what you're saying is right — the security guarantees of Chia are not the EXACT same as Bitcoin, but it's not clear to me that this means they're "weaker" for the reason you described.
> As soon as one honest farmer has added a foliage block, the foliage becomes impossible to re-org beyond that height with the same PoSpace
With ~350k Chia full nodes, the likelihood of colluding farmers being able to pull this off is a lot lower than if there were only the ~14k full nodes of Bitcoin, no? And Chia is still pretty new — it had a maximum of ~750k full nodes before.
I guess it's kind of a trade-off, huh? Bitcoin's big centralized pools are a worrying source of attack that's not there in Chia (the official pooling protocol has the farmers sign blocks so it's completely decentralized and pools using it can't orchestrate attacks). https://docs.chia.net/docs/11pooling/pooling
Every proof-of-work solution in Bitcoin commits to the transaction history, but the proofs-of-space in Chia only commit to the farmer public key.
Which is used to separately sign for a block of transactions. Which allows a collusion of recent farmers to produce alternate transaction histories at no cost.