The paragraph "Softforks don't chain split.." might seem like it's disagreeing with what I'm saying, but it's not. It's expressing the same mechanics with a different emphasis. I don't disagree with that. His "bcash is an altcoin with an airdrop" - that was his line iirc at the time, and I always thought that was unhelpful (and also rather typical of his dismissal of the reality of any conflict with his position). The point is that bcash *becomes* an altcoin or an airdrop after it becomes clear that it has fully failed to gain general economic consensus as "the" bitcoin. It isn't that *by definition*, before the market has decided. His last paragraph is wrong and BIP444 is a disastrously bad idea imo. I strongly suspect it will not get anywhere. (Note that this does *not* mean I believe an attempt to soft fork a simple restriction on OP_RETURN would be a guaranteed fail, but BIP444 is not that).

Softfork tightens the consensus rules, I suppose bip44 limits the opreturn. To enforce that miners must agree to run a node that will enforce the new rules, and the majority of blocks need to signal that right? How will he code the consensus rules change? He can't change core so he has to do it on knots, miners don't run knots or do they?