Marco 🎩

The most difficult part of this entire struggle is trying to convey information about a deeply technical debate to non-technical people. Statuette is undoubtedly one of the best at succeeding in conveying this. Yesterday, the network state team published an update to snapchain, which took effect two hours after release. This broke replication to any peer who did not upgrade. When I worked for merkle, we had a process for releases, so that non-emergency (which this was) updates gave node runners an entire month to upgrade. When neynar took over, with no announcement to change in process, they shortened it to a week window. And apparently, as of yesterday, it has been shortened further to two hours. This is fine when you're rolling out code to your own fleet. You control it! But when you roll it out, with no announcement, to a p2p network, you're telling people that their nodes don't matter. And that message is true anyway, because any node that isn't theirs is actually read only. When you hear someone posting messages to a snapchain node they own, it doesn't actually participate in any meaningful way – read only nodes simply forward on the message to the NS-controlled validators. Hypersnap will change this in the next release for our side of the fork – we will resume a more reasonable upgrade cadence, and we will cut ties with their validator set so that any node can actually participate in building data on the network. We aren't surprised to see what happened with the new (sort of) snapchain team, and our bar is so low for disappointment with their team that this is essentially a non-event. They embody neither the ideals of decentralization, nor the technical skills to achieve it.