> If we agree to include this in beta-2, I'm OK with proceeding, but I think we'll need a separate super-majority vote on ignoring the agreed beta rules, which might be harder than just re-rolling the release?
This position, on a release vote, by precedence sounds more like a -0. A valid concern has been raised, needs to be heard, but can and should be resolved within this thread and vote. It has my vote to be added to 4.0- beta2. I think it's worth pointing out that this ticket was never marked with a fix-version "4.0-alpha", so it was never visible as a blocker to the first beta release. Being strict here IMHO would mean reverting the commit and posting it to 4.1/5.0, rather blocking beta1. And… isn't the only thing that breaks API compatibility (and our beta rules) here is this one line: https://github.com/apache/cassandra/commit/d8993934e976d8edb94cbfe2974688dac63c5db5#diff-4805e34bd9553ede03778be66ddc06c7R268 I can cut a new 4.0-beta1, no problems there. The bar to re-cutting a proposed release should be low enough that it's an easy enough call to request. I will wait a few hours before doing so, to see what eventuates here…
