GraphQL is celebrating ever-increasing success in the API zeitgeist: a flexible language for describing precise, consumer-defined queries to data sources in a language capable of describing complex relationships between entities of any shape we can imagine. However, to a developer coming from the world of OData, it may seem quite limited at first: how do I filter my data or build complex aggregations? How do I avoid race conditions when I start changing it?

In this talk, we'll look at the OData specification, and discuss a number of its ideas that (I think) would be worth adopting into the culture of GraphQL. We'll look at the differences between the two ecosystems, the open source projects that try to reconcile those differences, and probably end up inventing a third competing standard along the way.