Via James I learned that earlier this month was released Natural Earth v2.0.0. A quick reminder: "Natural Earth is a public domain map dataset available at 1:10m, 1:50m, and 1:110 million scales. Featuring tightly integrated vector and raster data, with Natural Earth you can make a variety of visually pleasing, well-crafted maps with cartography or GIS software." We mentioned this dataset in the past quite a few times.
From the 2.0.0 release notes: "The 2.0.0 release focuses on 7 major areas and is available to download today à la carte at NaturalEarthData. ZIP combo downloads of all vectors: SHP (279 mb) or SQLite (222 mb) or QuickStart kit for ArcMap and QGIS (165 mb). [What's new:]
Via James I learned about Microsoft's launch of OpenGeocoder, a geocoding tool using bounding boxes for places. What's nice it that Microsoft gives back all resulting data to the public domain. There's a JSON API too.
From the about: "What is this? OpenGeocoder is an experiment in creating and serving geocodable results. Places are turned in to bounding boxes. Large datasets, processing and geocoding software is skipped. Instead a simple mapping between strings and boxes is used. All data submitted is placed in the public domain for anyone to use.
How do I use OpenGeocoder? Search using the text box. If your result is not found you are given the ability to add it. Drag the rectangle corners around until the rectangle covers the place you searched for and then click 'Save'. Your data is placed in to the public domain for anyone to use."
While we mentioned Geohash over three years ago at its launch, it's only last week that I attended a presentation on the subject and got excited by its potential. I'd defined it as a way to encode and index geographic coordinates at multiple scales with a single value. This may offer major benefits for some use cases. One of the drawbacks, not mentioned in the Wikipedia article (see below) is the size of these encoded coordinates for large datasets, but the presenter, Nouri Sabo, shared his tips for dramatically reducing it (up to about 80% in his use case).
The geohash.org website provides some information on what geohash is, but the Wikipedia article offers a much better overview and more details, from the introduction: "It is a hierarchical spatial data structure which subdivides space into buckets of grid shape. Geohashes offer properties like arbitrary precision and the possibility of gradually removing characters from the end of the code to reduce its size (and gradually lose precision). As a consequence of the gradual precision degradation, nearby places will often (but not always) present similar prefixes. On the other side, the longer a shared prefix is, the closer the two places are."
And selected passages:
The main usages of Geohashes are: