Better world-countries.json.

This was generated using the Natural Earth data set, which has nicer
simplification than my previous approach using MapShaper. Also, this includes
some updated boundaries, such as the split between Sudan and South Sudan. And,
counties now have ISO 3166-1 alpha-3 codes.

examples/data/README

@@ -1,5 +0,0 @@
-These are derived from the cartographic boundary files from the 2000 U.S. Census:
-
-http://www.census.gov/geo/www/cob/bdy_files.html
-
-Then, MapShaper was used to simplify the geometry, and ogr2ogr to convert the shapefiles to GeoJSON. Some additional work was done to preserve the FIPS codes, which are dropped from the shapefiles by MapShaper.

examples/data/README.md

@@ -0,0 +1,13 @@
+## World Boundaries
+
+These are derived from the public domain [Natural Earth](http://www.naturalearthdata.com/downloads/) cultural vector files, 110m resolution. Then, ogr2ogr was used to convert to GeoJSON. Lastly, the data was cleaned up slightly, removing extra properties and a degenerate edge from Antarctica.
+
+ collection.features.forEach(function(d, i) {
+   d.id = d.properties.ISO_A3;
+   d.properties = {name: d.properties.SOVEREIGNT};
+ });
+
+## United States Boundaries
+
+These are derived from the cartographic boundary files from the 2000 [U.S. Census](http://www.census.gov/geo/www/cob/bdy_files.html
+). Then, MapShaper was used to simplify the geometry, and ogr2ogr to convert the shapefiles to GeoJSON. Some additional work was done to preserve the FIPS codes, which are dropped from the shapefiles by MapShaper.