Codes are Not Identifiers

In a former post on how 'specific keys are the main cause of copying in databases' I demonstrated that codes are not identifiers for information entities and, by extension, they cannot be unique identifiers (UIDs).Why Do People Use Codes?How come, if codes are not identifiers they used so extensively in the company? Because, when used correctly, they are a very beneficial method of distinctively referring to real or abstract objects but, and this is vital to understand, they do NOT uniquely identify those objects.A code is just a letter, number, word or other symbol used to uniquely refer to something.In the example of the warehouse which I used in my previous post 'unique keys cause duplication in databases,' I used a five-digit numerical code (14563). Some would argue that this is the reason why identification failed. If I had used a more sensible code, then a structured code, then that identification would have been easy.If I had asked you, for instance, to pick up part number EP10KW then you'd have had no problem. Is this true? You could have used all of your creativity and tried to work out what the code meant. Is it perhaps an Electric Pump with 10 KiloWatt evaluation or EPoxy Resin at a 10 Kilo bath and White in color or an Extension Pole 10 meters in length created by Kite & Ward? There's absolutely no means of telling. You would be simply imagining.QUACKsWhen used correctly, structured codes, for example, EP10KW can be very useful. From the Integrated Modelling Method, they are referred to as QUACKs (Quick Unique Alternative Code or Key). They are not an identifier - merely a reference code.In order to be of any application, you have to get a means of knowing what the code identifies. This is ordinarily a listing or a catalog like the following.QUACK Name Power Source RatingEP10KW Pump Electric 10 Kilowatt EP20KW Pump Electric 20 Kilowatt PP05HP Pump Petrol Engine 5 horsepower