Tbh, the knots side looks more sus, and I was initially very pro-knots. I don't care about forks - we should do it more, not less. The case for core30 is that these larger op_returns are prunable. If I can prune nonmonetary data and run a node cheaply, then I have no reason to care if its expensive to run full nodes. Anyways, everyone is fighting the wrong battle. Segwit was dumb. 4mb is a large block. Fix **_that_**