EIP-3436: Expanded Clique Block Choice Rule

  1. Then choose the block whose validator had the least recent in-turn block assignment.
  2. Then choose the block with the lowest hash.

As for those rules… Rule number 3 means that determining whether to reorg is not really bounded. If it were 7 signers, only 5 were active, and chugged through 10M blocks. And suddenly number 6,7 pops up, and constructs a block each. Then we’d have to go through 10M blocks while searching for the ‘least recent’.

I guess I don’t see why we don’t just skip 3, and go directly to 4 and compare hashes ? That seems like the ultimate tie-breaker, and it’s highly ‘localized’ and cheap to perform.