MapBox now supports a cloud data library that is fully integrated into the new version of TileMill 0.4.2. This new geodata library is a collection of public datasets optimized for use with TileMill hosted on Amazon’s cloud infrastructure that users can browse directly from TileMill’s layer configuration interface. Selected files will be added directly to your project, downloading only the files you need once you need them.
The MapBox Geodata browser in TileMill 0.4.2
The library currently contains the full Natural Earth 1.3.0 release. This is a lot of data with plenty of helpful fields and variations for map styling. Land, water, and coastlines, national and sub-national administrative boundaries (both lines and areas), major highways and railways, populated places, urban areas… the list goes on. There are 59 files in all. Huge shoutout to Nathaniel Kelso for maintaining such an amazing public resource.
All the files have all been been optimized for TileMill so the files will render as fast as possible and you shouldn’t have to worry about things like projection settings. Details on these optimizations as noted in the manual:
All text within the data has been converted to UTF-8 which is the preferred encoding for working with data in TileMill. All the data has been reprojected to Google Mercator projection, and is properly auto-detected as such by TileMill. All geometry data has also been cropped (where necessary) to fit within the Google Mercator square Earth bounding box of 180.0° east/west and 85.05112877980659° north/south. All shapefiles have been indexed using Mapnik’s shapeindex tool. (The article Speed Optimization: Shapefile Indexes explains the benefits of this and how you can make sure your own shapefiles are indexed.)
We plan to continue to expand the data library with more high-quality free data sources, as well as include any updates that come from Natural Earth. We hope this integration makes it even easier for TileMill users to get designing awesome maps quickly. As always, the latest version of TileMill can be downloaded for free on TileMill.com, and the code is open source on GitHub.