In the next two chapters, we’ll venture into the side of PCG where we generate meshes from real-world Geographic Information System (GIS) data. In this chapter, we’ll write a Blender Python script that generates city meshes from building footprint data (a form of GIS data). We’ll start with an introduction to what building footprint is, what a dataset consists of, and how we can inspect and plot its contents to gain insight. After that, we’ll go over where and how to source building footprint data and how to edit them in the open source GIS software QGIS to convert them into the formats we can work with in Blender Python. Once we’re familiar with the composition and limitation of building footprints, we’ll formulate a design for a generator and start implementing it in Python. Using a top-down approach, we’ll compose a series of functions, each doing a unit of work at a finer granularity than the last. Finally, we’ll test the generator with two provided building footprint datasets and produce skyline meshes of Battery Park, NYC, and Havana, Cuba.