I'm addressing the question of whether we all had better assume all the RubyGems published after this incident were compromised, and my response is "that is probably not rational since the actor in this scenario had all this access legitimately just days beforehand". The rest, I don't care.