Films on the Map - OpenStreetMap-derived map data (ODbL 1.0)
=============================================================

Site: https://films-on-the-map.pages.dev
Attribution: (c) OpenStreetMap contributors - https://www.openstreetmap.org/copyright

This database is made available under the Open Database License:
https://opendatacommons.org/licenses/odbl/1-0/.
Any rights in individual contents of the database are licensed under the Database Contents License:
https://opendatacommons.org/licenses/dbcl/1-0/

Files
-----
citymaps/<slug>.json   The maps of the 8 city and region guides (london, new-york, northern-england, oxford, palermo, rome, scottish-highlands, taormina).
                       Layers: land (built from the OSM coastline), water, park, river, rail,
                       roads r0 (motorway, trunk), r1 (primary), r2 (secondary, tertiary, unclassified),
                       r4 (residential, living street, pedestrian - only inside the detail boxes),
                       and place labels (name and position from OSM place nodes).
uk-osm-layers.json     The OpenStreetMap layers of the Harry Potter map of Britain and Ireland:
                       "thames" (centre line of the River Thames, Oxford to the estuary),
                       "parks" (the large parks of central London) and "lochs" (the Scottish Highland lochs).
                       The land layer of that map comes from Natural Earth (public domain) and is not included.

Source and method
-----------------
Source: OpenStreetMap, downloaded through the Overpass API on 30-31 October 2026.
Processing (the site's build tools city_map.py and uk_map.py, using the shapely library):
- projection to a frame 1000 units wide (equirectangular, x scaled by the cosine of the frame's middle latitude);
- the land polygon is built by polygonising the coastline (land on the left of the line direction);
- lines of each layer are merged (linemerge) and areas are unioned;
- simplification with a tolerance of about max(4.5 m, 1.1 px of the view), then coordinates rounded;
- every layer is clipped per detail level: one coarse level for the whole frame and one level for each
  area view, with a small overlap so that fills have no seams;
- the result is stored as SVG path strings in map units, plus the projection parameters ("proj"/"frame").
The SVG maps on the site (/assets/maps/<slug>.svg and the inline maps) are Produced Works made from these files.

Questions: https://films-on-the-map.pages.dev/en/contact/