Overthinking numbers and strings

We model and build software with data constructed from too few basic types: typically including only one kind of number, and one kind of string. Modelling would be so much more interesting with all three kinds of numbers - ordinal and nominal numbers, as well as ‘normal’ cardinal numbers. Similarly, there are four kinds of string that we systematically mix them up, leading to further bugs and bad designs.

Thinking more deeply about numbers and strings reveals how overly primitive modelling causes both beginner mistakes and decisions we get stuck with for decades. Attendees will learn about house number and gender bugs, why we keep trying to organise personal names, and The Great Array Index Scandal. You will gain new perspectives on daily data types, awareness of recurring modelling bugs, and new inspiration for building better software.