Generator functions have been a part of Python for many years already, and are a well known technique for creating iterators. But generators have a few lesser-known aspects, including their “send” method and the “yield from” syntax. Many Python developers shy away from using them, unsure of what they would do, or how they would be useful — seeing coroutines as a solution looking for a problem. In this talk, I’ll tell you why coroutines can be useful, and how thinking about them as in-process “nanoservices” puts us in the right frame of mind to determine when they would and wouldn’t be appropriate. Among the topics we’ll explore are: Generator function basics next vs. send How to build a coroutine Thinking of coroutines as nanoservices Sub-coroutines The need for yield from Using yield from Ideas for general usage