Let's play with mixing partitioned, incremental, and versioned datasets. https://www.twitch.tv/waylonwalker