In KAFKA-391 I tried switching from (topic, partition) tuples to a TopicPartition case class in the producer/fetch request handling code. I think it helped quite a bit in terms of code clarity and access from java-api's. There are still lots of places where we have this:
https://gist.github.com/3718883 (this is after the edits to the request handling code and does not include other tuples that may be better written as case classes) This book [1] has a reasonable guideline (paraphrased): ".. tuples are almost too easy to use. Tuples are great when you combine data that has no meaning beyond an A and a B. However, whenever the combination has some meaning, or you want to add methods to the combination, it is better to go ahead and create a class... It makes your intentions explicit, which both clears up the code for human readers and gives the compiler and language opportunities to help you catch mistakes." I think we generally intend to follow the above but inadvertently go with tuples which subsequently spread all over. If there's no objection I can add the above to our coding convention page. Thanks, Joel [1] http://www.artima.com/shop/programming_in_scala_2ed
