In the past few weeks, the OGC (Open Geospatial Consortium) passed KML v2.2 as an OGC standard. This is a major step forward for the OGC and for Google, as it moves a key industry specification into the world of open standards.
KML is a very complementary to Geography Markup Language (GML) the other key encoding standard from the OGC. Where GML provides the mechanisms to describe geographic feature types, KML provides the mechanisms to visualize these features on a map or globe, and to control the user’s navigation over that map or globe. This complementary character is illustrated in Figure 1. Here GML data from a WFS is styled on the fly to KML and using a network link is made visible in Google Earth through the latter’s support for the OGC WMS interface. The coloured line segments in the diagram are from a dynamic link model of highway traffic using GML dynamic features and show the occupancy and rate of flow of vehicles (in real time) for the Toronto highway system. Figure 2. shows a more static picture of roadways near Red Deer Alberta that was part of a recent national CGDI demonstration. While Google Earth was used as the client in both of these examples, the adoption of KML as an open standard at the OGC, means that many other earth browsers and mapping clients will be usable in the very near future.
Figure 1. Real Time Traffic Data Styled from a WFS (GML) to KML
As a result of the passage of KML at the OGC we can now anticipate further developments not only within KML itself, but also in terms of other specifications such as the WMS (Web Map Service), Styled Layer Descriptor (SLD), and the Web Map Context specification. Consider for example, the Styled Layer Descriptor – the Symbology Encoding component in particular. This provides styling rules to transform feature data encoded in GML into a target visualization language. In the past this target visualization language was modeled on SVG, and the visualization description now in SLD is essentially derived from SVG (Scalable Vector Graphics). While SVG has some very strong graphics description capabilities, in some areas much stronger than KML, SVG is only 2D and does not incorporate concepts for navigation (e.g. LookAt in KML has no correlate in SVG, SVG graphics window while supporting images, has nothing corresponding to PhotoModel in KML). We can thus anticipate that SLD may be extended in the future to enable more use of KML capabilities.
Figure 2. GML data (from WFS) styled to KML for visual presentation through WMS in Google Earth
The development of KML has also influence the way OGC itself works, with a greater emphasis on running and adopted code prior to standardization. This reflects both the widespread adoption of KML, but also the greater maturity of the OGC and OGC Standards in general.
Future developments of KML are hard to predict but I would bet on enhancing some of the graphics functionality, and perhaps a stronger symbolization construct – maybe even modeled on SVG!
The ground work has now been laid for KML as a global standard for interactive map display and map navigation. Adoption outside of Google is already growing fast with support for KML in Microsoft Virtual Earth already announced and support in many other software products on the near term horizon, including automated validation of KML data streams.