Before then, people built distributed systems and (sometimes) assumed they didn't have failure modes. Now, people build distributed systems and assume they don

Before then, people built distributed systems and (sometimes) assumed they didn't have failure modes. Now, people build distributed systems and assume they don't have failure modes AND THEN post about how they've "refuted the CAP theorem". :)