The next big thing in standardized spatial data exchange format?

Question

I remember reading something a few months back, in the latter half of 2012, about a coordinated group effort to design a new standard for sharing geographic / spatial data. It has or aims to have the potential of replacing the showing-its-age (20+ yrs) de-facto standard of the Esri shapefile as the means of choice for widely sharing our data across all devices and platforms. Aside from addressing the most pressing shapefile limitations, this new thing incorporates raster and metadata too.

This big-important-thing I don't recall is not SpatiaLite - Shapefile of the future? or Are there any attempts to replace the shapefile? but some of the conversation swirling around it was similar. I spent quite some time today trying to track this down, but my search-fu proved not up to the task (which might not bode well for its future).

Please, what is the name of this elusive project that has the potential to deeply impact all our workflows? (Assuming it gets off the drawing board, lord knows we need something to.) And, how or where does one get involved or at least keep track of it?

Answer 1

The idea you're probably thinking of is GeoPackage.

Background.

The background to the requirement came from the US Army Geospatial Center which is part of the Corps of Engineers, who were looking for a way to put maps and other information (points or interest, routes, photographs and other observations) into a simple "one file has it all" format for mobile applications. The original development was done on public mailing lists (one for vectors and one for rasters/tiles) on Google Groups.

When the OGC started the OWS-9 testbed activity, the development and prototyping moved into the OGC. The reasons for it are a little complex, but it does appear to have been done with good intentions. The only problem was that you had to be "in" the OGC world to see what was happening. The two Google Groups lists are basically dead.

As OWS-9 came towards the end, the OGC (prompted by the NGA and US Army Geospatial Center) spun up the standards working group (SWG) to develop a "proper" spec. The OGC has more detail on this.

The OGC specification development process was initially overseen by Paul Daisey. It was an interesting process in that it attracted a lot of external attention. Initial work was done in Microsoft Word, but it switched to development on GitHub. As of 2017, the SWG chair is Jeff Yutzler. Note that the version on GitHub as markdown is not an official spec - its a work in progress, much like source code as it moves from official release to official release.

Technical

The specification is somewhat difficult to read (but hopefully hard to get wrong in the implementation stage). It tries to be unambiguous, not nice. The best way to understand what is happening is to find the diagram showing the table relationships. Stare at that for a while, and it might start to make some more sense. Then look at some sample data (note: may not be current to the latest spec).

Implementation

There are a range of implementations (see the http://www.geopackage.org/ web site for links). Disclaimer: I worked on some of them.

Future

As of late 2017, development of the core specification has slowed - there are new versions, but differences are mostly pretty small. There is a fairly open extension mechanism - see http://www.geopackage.org/extensions.html for an incomplete list. The elevation extension was the first one that the SWG worked, and the near future work is on related tables, which is a way to link a geospatial feature with some other data (e.g. a photo of a landmark, or PDFs of the take-off and landing "approach plates" for an airport).

OGC is sponsoring some member organisations (companies, universities) to do other work in its testbed activities.