Thank you very much, Owen, for making crystal clear that the functor I is simply a more complete way to specify the whole dataset Gr(I), as it also includes how the data relates to our schema C.

I've learned something, and I'll mention the _Grothendieck construction_ next time I relate data with a schema!

Also, I've just decided to enumerate my puzzles per lecture, simply mentioning the lecture number to make things abundantly clear. So that should have been puzzle rf38.1.