kdb also allows you to do much more than SQL, since a select is more or less just syntactic sugar for a certain set of operations on columns as arrays. The full language is available in the context of the query, or you could just treat your table as a bunch of arrays in the context of a larger program.
It's a really elegant way of dealing with large amounts of data, although the downside is that you've typically got to build a lot of the nice-to-have dbms type infrastructure yourself.
It would be interesting if BigQuery or Redshift ever figure out that they could have a much more powerful system if they stuck an array language on the front of their storage engines.
It's a really elegant way of dealing with large amounts of data, although the downside is that you've typically got to build a lot of the nice-to-have dbms type infrastructure yourself.
It would be interesting if BigQuery or Redshift ever figure out that they could have a much more powerful system if they stuck an array language on the front of their storage engines.