"It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts."
— Sherlock Holmes (Arthur Conan Doyle), A Scandal in Bohemia
In data warehousing and business intelligence, there are many approaches to data modeling. We hear of personalities such as Bill Inmon and Ralph Kimball. We talk of normalization and dimensional modeling. But we also might have heard about how QlikView can cut across all of this—we don't need to worry about data warehousing; we just load in all the data from source systems and start clicking. Right?
Well, that might be right if you want to load just a very quick application directly from the data source and aren't too worried about performance or maintainability. However, the dynamic nature of the QlikView script does not mean that we should throw out all of the best practices in data warehouse design that have been established...