All files are static--there is no real backend server or database that stores the
data. All data on example setups comes from src/examples/*. Update those
examples, then run npm run generate to parse it into the files in
dist/examples/. The files here are broken up by wheel size and offset, which
is easier for the web interface to query (eg: it can just GET /examples/<car>/<width>/<offset>.json).
This might not scale well, but GitHub pages is free :D