Geospatial Standards Move Toward Implementation and Interoperability

Cliff Kottman Vice President, Open GIS Consortium
Friday, 30 April 1999 UTC
Information technology standards are often book-sized documents of rather arcane technical information and directives. They are dry and brittle. Yet, the best information technology standards directly affect our daily lives, change the way we do business, and, indeed, change the world. This is the case with emerging geomatics and geographic information standards. The fact is that many daily tasks and activities can and will be made easier, more pleasant, and more professional using standards that are emerging from ISO/TC211, and being implemented by the members of the Open GIS Consortium. The Open GIS Consortium (OGC) is a not-for-profit global trade association, with about 180 members and growing, dedicated to open geodata and geoprocessing in distributed and heterogeneous environments. (See: ). OGC is now working in close cooperation with ISO to achieve a family of interoperable implementations of the various standards being forwarded by ISO/TC211. The reason OGC exists is to address the stovepipe nature of today's geodata and geoprocessing environment, and the constraints this vertical architecture puts on geo-enabling today's enterprise environments. New algorithms and best methods for travel, positioning, advertising, planning, resource management, facility management, and similar disciplines cannot easily be shared. Instead, they are trapped inside stovepipes, rendered vendor-specific, and unavailable for use by wider audiences. Even within a single vendor's environment, it can be extremely difficult or virtually impossible to invoke remote processes in heterogeneous distributed computing environments within a single session. It is the vision of OGC to achieve industry-wide consensus on the fundamental classes and interfaces upon which Geographic Information Systems are constructed. By carefully introducing families of these classes and interfaces in the proper order, the Consortium enable services, applications, and user environments to be implemented. The actual implementations, that is, the software that enables data and services to be invoked, and to behave in the way proscribed by the interfaces, is the duty of the industry building content and applications. The implementations distinguish themselves in the marketplace and, through the new and improved goods and services they supply, generate a larger geospatial economy for all of us to enjoy. At the ISO/TC 211 Plenary meeting in Vienna, 1999-03-05 a co-operative agreement between ISO/TC 211 and OGC was accepted. Its first implementation was at the OGC meetings in Enschede, The Netherlands, on 1999-04-20. The result will be tighter linkages for the family of information technology standards related to geography, mapping, land use, land ownership, surveying, and so on. The fundamental goals of the cooperation between ISO/TC211 and OGC are to ensure conformance to the ISO standards in commercial implementations, and to achieve interoperability across disciplines and user environments. The idea is that geospatial information should move as smoothly between GIS applications as word processing information does, and that the brand name of new components for your GIS should matter as much as the name of your long distance telephone carrier: that is, not at all. Interoperability never happens by accident. It takes careful planning, a consensus process, and representation by all of the communities in which interoperability is desired. OGC provides exactly this mix of government, academic, industrial, and trade discipline representation. The Cooperative Agreement between OGC and ISO TC211 is based upon a Terms of Reference for the Coordination Group created by the Agreement. These terms call for a continuous analysis of the opportunities whereby ISO and OGC can cooperate, to call these to the attention of both bodies, and to ensure that proper experts of each body are engaged in the activities. The Coordination Group makes no technical recommendations itself, but is strictly attending to the question: are the right people addressing the important questions at the appropriate time? The Coordination Group Convenor is John Rowley of the UK. A secondary goal of the Cooperative Agreement is the advancing of OGC Implementation Specifications toward International Standard Status. Implementation Specifications are documents written for software developers. They contain directions that are so explicit and detailed, that if they are implemented (that is, turned into working software) by two different developers, the implementations should "plug and play" for each other. The first tests for interoperability are now in the planning stage. In addition, OGC has developed a test environment, and has already tested two implementations for conformance to the specifications. (Note that testing for conformance to the specifications is less rigorous than testing for interoperability.) OGC has recently submitted the OGC Simple Feature Implementation Specification, SQL Option, (version 1.1) as a support document for an ISO New Work Item Proposal (NWIP) centered on Simple Feature Implementation Specifications. "Simple Features" in OGC are features whose geometry may be represented by points, lines, and/or polygons. Several additional NWIPs are in work and will be submitted from OGC to ISO/TC211 soon. The OGC Simple Feature Implementation Specification has three options: one is based on CORBA (the Common Object Request Broker Architecture), one on COM-based technology (using the Component Object Model of Microsoft), and one is based on SQL. All three are freely available at the OGC web site: OGC is harmonizing the three options, and has engaged GEOBASE (UK) to make these specifications conformant to the ISO template for International Standards. OGC intends to maintain all future versions of the SQL Option in ISO-like format from that point onwards. Significant technical advances are expected to occur at the next few OGC meetings. Meetings are scheduled every two months. The next four meetings are June 14-17, 1999 Reston, Virginia, USA August 9-12, 1999 Southampton, United Kingdom October 3-8, 1999 Tokyo, Japan December 6-9, 1999 Los Angeles, California, USA Among the technologies to be advanced at those meetings are Catalog Services, Gridded Coverages, Coordinates and Transformations, and Feature ID and Relationship technology. Additionally, a testbed is being developed in Gaithersburg, Maryland, to advance the interoperability of presentation technology and Web access channels for geodata. The work on Catalogs will address the browsing and publishing behaviors of clients and servers supporting catalog services. Of special interest is the bridging of message-based protocols such as http and Z39.50 with object invocation protocols such as CORBA, so that heterogeneous mixes of clients and servers of both technologies are enabled to support your queries. The work on Gridded Coverages will provide consensus implementation specifications for fundamental behaviors that Gridded Coverages need to support. Gridded Coverages include digital elevation matrices, orthoimages, digital images, and other rectangular arrays of sample values. The fundamental behaviors will provide the abilities to evaluate the array at a given point, to georeference the grid, to convey the semantics of the grid values, and to manipulate many Gridded Coverages (for example in paneling and mosaicking.) The testbed in Gaithersburg is perhaps the most interesting OGC development. Here, a half-dozen sponsors have challenged industry to craft solutions to demanding scenarios based on Disaster Management and Command and Control situations. More than a dozen technology providers have volunteered to participate on a cost-sharing basis to stand up various parts of an overall solution. The end product will be demonstrated this Summer: an extremely flexible architecture of geodata, data servers, style generators, mapping servers, service registries, and browsing and publishing clients. From this testbed, interfaces that work well will be identified, abstracted to their specifications, and submitted to the OGC consensus process. This is an exciting time for Open GIS. The foundations are being laid for a large and beautiful new architecture of GIS components. -- end --