SceneCity has its own socket types and they have nothing to do with Blender’s sockets. The usual color codes you’re used to (yellow for images, blue for vectors and coordinates, grey for values etc…) don’t have the same meaning in a SceneCity node graph.
You should not connect a socket type to another socket of a different type. If you need to convert a socket type to another, you may use a converter node, if such node exists.
Mesh data is SceneCity’s internal representation of a mesh. It contains the vertices, edges, faces, normals, UVs etc… Not to be confused with a Blender mesh.
A map is like a grayscale image, it has only one channel, no color. But unlike an image, it is not a raster, it is continuous. It may be based on a procedural noise function, or based on a raster image, or mixed.
A texture set is like a PBR material, it’s a collection of images, each one for a different channel in the PBR workflow (and other useful maps). Ultimately they are meant to be output as images and used in materials / shaders.
A collection of Blender objects.
Not to be confused with SC objects.
Paths are 3d polylines, they are a sequence of linked points, with a start point, and an end point. Unlike graphs, they cannot branch, meaning each point along the polyline can only have 2 links at most: one before, one after. A point cannot be disconnected, except if the path has only one point.
These are similar in concept to Blender’s bezier curves. They are in 3d.
A collection of the internal representation of objects in SceneCity. They can later be converted back to different types of Blender objects using specific nodes, for instance the objects instancer node.
Not to be confused with Blender objects.
A collection of buildings. A building is made of SC objects, and has a size in grid cells. Typically placed on a grid.
A collection of road portions. A road portion is made of SC objects, and has a type: straight, T-crossing, X-crossing… Typically placed on a grid.
A grid is a 2d array. And in each cell can be stored any number of key / value pairs. For more details about grids, check the grid node. Can be used for lots of things, but typically used to place buildings and road portions.
Combines a map, with a physical size on X, Y and Z, so that the map has a scale in the 3d world. The physical size on X and Y is the exact size of the terrain on both axes, whereas the Z value is simply an indicator of the intended height where the map is equal to 1. Z is multiplied by the map values, normally between 0 and 1, but they can be any value, so the terrain heights can go beyond the height given in Z, and below zero as well.
Generic 2d geometries: points, lines, polygons.