For the past month or so, I’ve been working on implementing a pure-Racket client for Apache Kafka. Yesterday, it reached a point where it can do the bare minimum you would expect from it: produce data and join consumer groups to consume data. Kafka has a fairly large feature-set so there’s a ton left to do, but I figure this is a good time to announce the library and get feedback. If you’re interested in using Kafka with Racket, please give it a try and let me know what you think.

You can find the source code on GitHub and the docs on the Racket Package server.

PS: While writing the library, I needed an easy way to write lots of little wire procotol parsers, so I stacked my yaks and wrote a binary format parser generator called binfmt (docs).