Faking it - a UK reality (real??) TV show where people are given a crash course in some skill completely alien to their normal life (such a classical musician to dance DJ, or librarian to bodyguard) and then sent to compete against real exponents of the trade for a panel of experts to spot the Sexton Blake. The show is of course about the physical and mental pain experienced by the trainee as they are thrown from a building or whatever and nothing about learning a useful skill.
Flaking it - a term (probably of my invention) where we transform a traditional star schema (Kimball) model to something with snowflake characteristics - perhaps a Star-Flake.
Consider the product dimension, in a lot of traditional data warehouses this is represented by a fully denormalised product table that contains not only the attributes of the product (and possible useful compound attributes such as long name or name + product code) but the attributes of the parents and grand parents. For example take a box of washing powder, this could belong to a category of laundry products and that in turn could be part of grocery (non-food). All of this is represented in single table looking something like this
But is that the most appropriate way to represent a product? Some attributes of a product relate to storage and logistics (weight, storage temperature, pack dimensions, shelf-life) others relate to supplier performance (perhaps delivery lead time, order fulfillment targets) and yet other relate to product cost and other financial data. In these days where good governance demands that business roles are separated there is some sense in making product look more like this; a central core of descriptive attributes joined to a cluster of attribute tables each containing a single set of subject related attributes. This simplifies keeping information private and does not greatly hinder rollups on the dimension after all queries such as what is the total sales in May for all items with a storage temperature of -15C is not that meaningful and nor is what is our total spatial volume of laundry products sold in the year-to-date - possibly interesting but hardly actionable (what could you change if you knew it?)
