https://www.confluent.io/blog/publishing-apache-kafka-new-york-times/ https://www.confluent.io/blog/publishing-apache-kafka-new-yo... ^ this. All they needed

https://www.confluent.io/blog/publishing-apache-kafka-new-york-times/ https://www.confluent.io/blog/publishing-apache-kafka-new-yo... ^ this. All they needed was a database, or possibly a DB that supports row signing. I mean actually they could have done it with git. They don't publish that many stories an hour. Everything about this setup is just plain wrong, and to then boast about it, absolute madness.