Interesting that all of the successors to MapReduce fix all the major criticism DeWitt and Stonebraker have. (I'm thinking BigQuery, Snowflake, Spark) So the le

Interesting that all of the successors to MapReduce fix all the major criticism DeWitt and Stonebraker have. (I'm thinking BigQuery, Snowflake, Spark) So the lessons were re-learned and features re-introduced, but the massively parallel execution has persisted. Perhaps MapReduce was non-novel and flawed, but it certainly seems to have led to a flowering of rich, large scale data querying systems.