Location data is everywhere. From huge government databases of geographic features to your pictures in Facebook, it seems like almost every piece of information around nowadays is tagged with its location. However, it still seems that no one is effectively sharing information, or building the smart, next-generation systems that will surely rely on data from multiple, linked information sources.

We in the geospatial profession believe that location is the great common denominator. It has the best potential to be the bridge between systems of related data sets. But how do we devise a simple way to describe places and relationships between them that will appeal to 85% of the developer community?

The W3C Points of Interest Working Group has been tackling this problem throughout 2011, and is nearing completion of a fairly final draft of a specification. You can see the work on the POI Wiki, and join the public mailing list by sending an email here. In this group, we’ve created a relatively simple data model, and expect that people write POI data in JSON, RDF and/or XML format. The jury is still out on which format will win.

What I hope a common format will do is allow everyone, from Yelp to Facebook to humanitarian organizations and event defense departments all over the world, to share basic location information about common places. I think this will not only strengthen the core business propositions of these groups, but even enhance them, freeing up time from the mundane, repetitive task of maintaining accurate locations and creating more time for real application enhancements.

I believe this effort is the most important activity in the geospatial field at the moment, and will be writing and coding heavily around POIs. Join me and make 2012 the year of the POI!

Geography Markup Language (GML) 3.3 is now out for public comment. Not to say that this isn’t exciting on its own merits, but there’s an incredibly new feature in it — compact encodings! Clause 7 (page 16 in the PDF) has a great little section called “Compact Encodings of Commonly Used GML Geometries.” It finally defines a concise encoding for describing simple points, lines and polygons (and a few other shapes as well). For example, here’s a picture of the lineage of the new <SimplePolygon>, which compresses the old <Polygon> from 7 elements to 2.

In short, here are example points, lines and polygons that are now legal GML.

<Point>
   <posList>42.3 -71.9</posList>
</Point>

<LineString>
   <posList>42.3 -71.9 42.4 -71.8</posList>
</LineString>

<SimplePolygon>
   <posList>42.3 -71.9 42.5 -72.2 42.4 -72.1 42.4 -72.0</posList>
</SimplePolygon>

Download the standard here.

This from https://bugreport.apple.com when trying to log in and submit a bug.

So the bug reporting system has too many bugs to allow bug reporting?

Namespaces can get near on impossible to process sanely.

http://blog.technologyofcontent.com/2010/01/json-vs-xml/

see the “Against XML namespaces” section

The “Internet of Places” is a nice article putting forth an ecosystem of information and services to bring the geospatial world better in line with the Web. Or one could even say it suggests an information ecosystem where geospatial data would help the Web make a quantum leap in information linking and tagging.

I proposed a similar architecture in my AAG presentation earlier this year called, Historical Gazetteers and Points of Interest. You can also find it on OGC Network.