3 ms·
I'm not sure what you mean. Isn't the typical fork selection method for proof-of-stake to choose the chain with the most validators?
by shawnz 5y ago
I'm not sure what you mean. Isn't the typical fork selection method for proof-of-stake to choose the chain with the most validators?
- wmf 5y agoUnfortunately it's more complex than that. You can't rely on number of stakers for anything because of sybil attacks. You also can't rely on amount of stake because some of it could be double-staked (you can detect this but it still may not help you choose which fork is legit). The solutions to PoS fork choice don't really fit in an HN comment.
- shawnz 5y agoI'm not trying to say it's a simple problem, but I think "it's a glorified Git" is a mischaracterization