Added Tables support, was very straightforward: DBCollections.jl/src/DBCollections.jl at 00000000be784dfad350b90e8355f74655f19737 · JuliaAPlavin/DBCollections.jl · GitHub!
I tend to agree with you and @pdeffebach that it’s generally useful, at the very least for IO. Like, pass a DBCollection to CSV.write directly.
Still – maybe it’s just my unfamiliarity with dataframes/tableoperations, but I don’t see how this Tables interface for DBCollections can help with data processing. As I understand, applying TableOperations here would just load all the data into memory before any actual operations.