I have worked with maps quite a bit, mostly for building Land Explorer (LX) and the Coop World Map (CWM), and then other smaller maps in a capacity as a freelance developer (PINF map, Strike Map)

I wouldn't describe it as particularly complicated, but it's helpful to know the various ways you can get a map in front of someone.

There's a difference between the software that displays the map and the source of your tiles, i.e. the thing that people load at different zoom levels of the map. So like on LX, you can use Mapbox as your web library and then tiles from Ordnance Survey, or on CWM we used MapLibre and MapTiler for the tiles. Google Maps gets you both. Usually the choice comes down to the relationship you want with your map and tile provider.

And then working with the data, there's a command line tool called ogr2ogr that will do most transformations that you could want. People make a dataset available in lots of different formats and I usually wanted to get a geojson from them so I could easily work with it in a React site for instance.

And then we used MySQL a lot because the database had built in polygon transformations so it would return queries like "which polygons are in this box" relatively quickly.

There's little things with working with maps that can be annoying, like working with Markers vs Layers and clustering and spiderfying and optimising for load time or performance etc.

I will say that staring at a map of the UK for 7 years off and on has really improved my innate knowledge of where things are.