Damn Apache Spark. Instead of supporting just one language that works perfectly well, let's support several languages that give users different results! It's so much fun to combine chunks of each language to get a fully working solution!

The worst thing is when I've seen people convert a DF from pyspark to scala spark because they need to send it to/ receive from a JDBC connection. They convert it on the way in and then on the way out again.

There are also 2 versions of Spark supported in R - sparkR and sparklyr. One is way behind the other and both are way behind python and scala.

Hey you gotta give it to them, at least the idea is ... okay

Haha, web developers ☕️