The story of reinforcement-learning-with-verifiable-reward-rlvr