Interesting paper: perhaps consensus isn't necessary?

Like Nano, every user maintains their own blockchain. That's where the similarities end though. What's interesting about this is it solves the ledger bloat issue, and can process 20000TPS, and is low latency, without requiring consensus via PoW or PoS. This would be a radical departure from Nano, but given the recent spam attacks, I'd believe this would give us a lot more headroom.

edit: another interesting paper that cites the previous: https://research.fb.com/wp-content/uploads/2020/10/FastPay-High-Performance-Byzantine-Fault-Tolerant-Settlement.pdf

edit 2: seems this is for permissioned blockchains...? But in appendix a, it mentions how to add / remove replicas, so not sure if this is a problem.

I haven't read the full paper yet (only the abstract), but isn't this how Nano worked at release, and similar to what's described in the original whitepaper? Voting only happened if a fork was detected

The problem was that this became a UX nightmare, as you never knew how long to wait for sure, and you could end up with long chains of transactions being "rolled back" after one part of the network detected a fork. It's much simpler to get confirmation up front and make sure everyone's on the same page

1 Like

Like Qwazhi said if you dont have consensus you dont have finality. How long would you wait until you confirm a payment? A fork could show up later rerolling the payment.

Block cementing is only possible because every block gets voted on.

5 Likes