In the note on page 388 you mention that each fetch request requires a round trip to the persistent store. But on page 385 you mention that the fetch returns objects based on their in-memory state in managed object context, including evaluating predicates (unless objects have not yet been loaded, e.g., from a previous fetch). Just a little confused about these two statements. Are you saying in the note on page 388 that all new fetch requests requires a round trip to the persistent store? Or does each fetch round-trip every time, e.g., to determine if any new/additional objects need to be pulled into the context to satisfy the fetch request, but (referring to 385) if the object is already in the context then the fetch will use that in-memory state vs. the persistent state of that specific object? Could you please provide some clarification? Thank you

I believe you have answered my question on page 397. Thanks