Node governance is bad, actually. It inverts the old IETF "running code and rough consensus" to insist on consensus before code. Which means all discussions are hypothetical, end up addressing problems that may or may not actually exist.

If you insist on running code first it gives you a concrete thing to build discussions on. "It might have performance problems!" great, let's test that hypothesis. It helps end the impasse of governance via filibuster, by requiring that everyone TEST their ideas and objections.

(Yes, I'm grumbling about modules again. I vaguely regret dropping out of the reformed modules committee, but its hard to have one's role to be to constantly say no forever. Given where things went, that's apparently what I would have been doing.)

(The _fact_ that whole process was reset, rather than engaging with what was in front of people was actually at the heart of the organizational failing anyway.)