After 7 years at Taboola (and 6 years at Perfectmarket before that), the time has come to change things a bit and move to a new company. “I open at the close” is what is written on a small ball that hides a life-restoring magic stone, and Harry Potter discovers this stone just being killed by YouKnowWho. Thanks to the magic stone Harry lives, and the story continues.
A useful trick: to be able to compile code at runtime and use its result(s) in a scala/java class!
In Sept 2020 I decided that I needed a home server to do some development. For the purpose I wanted a system able to execute a decently-sized Apache Spark job, and run some middle-tier TensorFlow model. The server is currently dual-booting Ubuntu and Win10.
In Java there are generally 2 options for a thread-safe Map:
Have an issue with moving around an odd class in an Apache Spark Dataset? Here is a practical solution!
I’ll be using github pages to host a simple static website. Seems to be pretty simple and handy to use :)