Everyone has seen "numbers every programmer should know" < https://gist.github.com/jboner/2841832 > If you're going to do complex data analysis, e.g. machine learning, you want your data access latency to be on the short end of this chart :) When you end up on the long end (like in RDBMS) this is known as N+1 problem.
But modern size data doesn't fit into memory. Distributed systems necessarily add latency, and to fix that we add caching, which hurts consistency. I blew up this thread further down about how Datomic's core idea is to provide consistent data to your code; which is the opposite of how most DBMS (including kbd) make you bring the code into the database.
Everyone has seen "numbers every programmer should know" < https://gist.github.com/jboner/2841832 > If you're going to do complex data analysis, e.g. machine learning, you want your data access latency to be on the short end of this chart :) When you end up on the long end (like in RDBMS) this is known as N+1 problem.
But modern size data doesn't fit into memory. Distributed systems necessarily add latency, and to fix that we add caching, which hurts consistency. I blew up this thread further down about how Datomic's core idea is to provide consistent data to your code; which is the opposite of how most DBMS (including kbd) make you bring the code into the database.