So the 200Gb loading and digesting part is totally separate from the Observable Framework, right? You just do it with a standard ( non wasm duckdb as part of ETL) and later you just direct Observable Framework to read and plot the 8Gb file? Thanks