Working on this a little more and the conclusion is to use conventional JOINs.
Just one more question, does the above query add any extra overhead compared to conventional JOINs? lets say an author would always have an outdoors book and a cooking book, would it be good to use the above query?
Its a pure hypothetical query, as in there are no books. Just playing around, I have your book but my joins keep giving me what I don’t want, That query gives me accurate results but only if the author is in both tables.
What kind of join does this FROM booksoutdoors, bookscooking do?
outdoors and cooking are possible values for a book category field, so one table for book and another for book-categories, and this allows a book to be both for outdoors and cooking.
book-categories {isdn, category}
book {isdn, some other fields,}
I’ll leave you to work out the authors bit…
but it’s not in the book table.