The corresponding DB schema
The most important points are:
All things on the map are entities (saved in the entities table), the database schema is built around that idea. Entities are timeless, they simly "are". There's nothing more to it. However, entities are "containers" for properties, an entitiy without a property is void and doesn't make sense.
There can be many properties (saved in the properties table) for each entity, the most obvious is geom which defines the shape of an entity on the map. All properties have a date (e.g. because for that date we have a source telling us that the entity had that property at that time), but also a computed start and an end date (i.e. the system detects that another property is valid after or before a certain date and not the one at hand). Other examples for properties could be: Height, Owner, Borders (e.g. if entity is a country), etc.
Entities can be related to each other (saved in related_entities).
You can have a look at the database schema in doc/dbanalysis/index.html
How to generate schema analysis
java -jar schemaSpy_5.0.0.jar -t pgsql -db vtm_dev -u -p -o data -host dhlabpc3.epfl.ch -dp postgresql-9.4-1201.jdbc41.jar -schemas "vtm,data_external"
This saves the schema into the data/ directory of the working directory. I then moved this data/ directory to doc/dbanalysis in this project.