As someone coming from the analytics perspective, ...
# random
t
As someone coming from the analytics perspective, I thought it was really bizarre having these SQL-like layers for a technology that was created partially due to "SQL being too hard". Of course, scalability and flexiblity was the main ambition. But it still was weird having a SQL interface for Apache Spark, and data science acting all embracing of "big data" solutions even though it made analytics absolute hell.