Editing Geography Markup Language (section)

==History==

===Initial work – to OGC recommendation paper===
Ron Lake started work on GML in the fall of 1998, following earlier work on [[XML]] encodings for radio broadcasting.  Lake presented his early ideas to an [[Open Geospatial Consortium|OGC]] meeting in Atlanta, Georgia, in February 1999, under the title xGML. This introduced the idea of a GeoDOM, and the notion of Geographic Styling Language (GSL) based on [[Extensible Stylesheet Language|XSL]].  Akifumi Nakai of NTT Data also presented at the same meeting on work partly underway at NTT Data on an XML encoding called G-XML, which was targeted at location–based services.<ref>{{ cite web|url=http://www.dpc.jipdec.or.jp/gxml/contents-e/history.htm|title=G-XML|url-status=dead|archive-url=https://web.archive.org/web/20091217094736/http://www.dpc.jipdec.or.jp/gxml/contents-e/history.htm|archive-date=2009-12-17}}</ref>  In April 1999, Galdos created the XBed team (with CubeWerx, [[Oracle Corporation]], [[MapInfo Corporation]], NTT Data, [[Mitsubishi]], and Compusult as subcontractors).  Xbed was focused on the use of XML for geospatial.  This led to the creation of SFXML (Simple Features XML) with input from Galdos, US Census, and NTT Data.  Galdos demonstrated an early map style engine pulling data from an Oracle-based "GML" data server (precursor of the WFS) at the first  OGC Web Map Test Bed in September 1999.  In October 1999, Galdos Systems rewrote the SFXML draft document into a Request for Comment, and changed the name of the language to GML (Geography Markup Language). This document introduced several key ideas that became the foundation of GML, including the 1) Object-Property-Value rule, 2) Remote properties (via rdf:resource), and 3) the decision to use application schemas rather than a set of static schemas. The paper also proposed that the language be based on the [[Resource Description Framework]] (RDF) rather than on the DTDs used to that point. These issues, including the use of RDF, were hotly debated within the OGC community during 1999 and 2000, with the result that the final GML Recommendation Paper contained  three GML profiles – two based on [[Document Type Definition|DTD]], and one on RDF – with one of the DTD's using a static schema approach. This passed as a Recommendation Paper at the OGC in May 2000.<ref>{{Cite web|url=http://portal.opengeospatial.org/files/?artifact_id=13252|title=GML in JPEG 2000 for Geographic Imagery (GMLJP2) Encoding Specification}}</ref>

===Moving to XML Schema – Version 2.===
Even before the passage of the Recommendation Paper at the OGC, Galdos had started work on an [[XML Schema (W3C)|XML Schema]] version of GML, replacing the rdf:resource scheme for remote references with the use of xlink:href, and developing specific patterns (e.g. Barbarians at the Gate) for handling extensions for complex structures like feature collections.  Much of the XML Schema design work was done by Mr. Richard Martell of Galdos who served as the document editor and who was mainly responsible for the translation of the basic GML model into an XML Schema.  Other important inputs in this time frame came from Simon Cox (CSIRO Australia), Paul Daisey (US Census), David Burggraf (Galdos), and Adrian Cuthbert (Laser-Scan).  The US Army Corps of Engineers (particularly Jeff Harrison) were quite supportive of the development of GML.  The US Army Corps of Engineers sponsored the “USL Pilot” project, which was very helpful in exploring the utility of linking and styling concepts in the GML specification, with important work being done by Monie (Ionic) and Xia Li (Galdos).  The XML Schema specification draft was submitted by Galdos and was approved for public distribution in December 2000.  It became a Recommendation Paper in February 2001 and an Adopted Specification in May of the same year.  This version (V2.0) eliminated the “profiles” from version 1. and established the key principles, as outlined in the original Galdos submission, as the basis of GML.

===GML and G-XML (Japan)===
As these events were unfolding, work was continuing in parallel in Japan on G-XML under the auspices of the Japanese  Database Promotion Center under the direction of Mr. Shige Kawano.  G-XML and GML differed in several important respects. Targeted at LBS applications, G-XML employed many concrete geographic objects (e.g. Mover, POI), while GML provided a very limited concrete set and built more complex objects by the use of application schemas.   At this point in time, G-XML was still written using a DTD, while GML had already transitioned to an XML Schema.  On the one hand G-XML required the use of many fundamental constructs not at the time in the GML lexicon, including temporality, spatial references by identifiers, objects having histories, and the concept of topology-based styling.  GML, on the other hand, offered a limited set of primitives (geometry, feature) and a recipe to construct user defined object (feature) types.

A set of meetings held in Tokyo in January 2001, and involving Ron Lake (Galdos), Richard Martell (Galdos), OGC Staff (Kurt Buehler, David Schell), Mr. Shige Kawano (DPC), Mr. Akifumi Nakai (NTT Data) and Dr. Shimada (Hitachi CRL) led to the signing of an MOU between DPC and OGC by which OGC would endeavour to inject the fundamental elements required to support G-XML into GML, thus enabling G-XML to be written as a GML application schema.  This resulted in many new types entering GML's core object list, including observations, dynamic features, temporal objects, default styles, topology, and viewpoints.  Much of the work was conducted by Galdos under contract to NTT Data.  This laid the foundation for GML 3, although a significant new development occurred in this time frame, namely the intersection of the OGC and [[ISO/TC 211]].

===Towards ISO – GML 3.0 broadens the scope of GML===
While a basic coding existed for most of the new objects introduced by the GML/G-XML agreement, and for some introduced by Galdos within the [[Open Geospatial Consortium|OGC process]] (notably coverages), it soon became apparent that few of these encodings were compliant with the abstract specifications developed by the ISO TC/211, specifications which were increasingly becoming the basis for all OGC specifications.  GML geometry, for example, had been based on an earlier and only partly documented geometry model (Simple Features Geometry) and this was insufficient to support the more extensive and complex geometries described in TC/211.  The management of GML development was also altered in this time frame with the participation of many more individuals. Significant contributions in this time frame were made by Milan Trninic (Galdos) (default styles, CRS), Ron Lake (Galdos) (Observations), Richard Martell (Galdos) (dynamic features).

On June 12, 2002, Mr. Ron Lake was recognized by the OGC for his work in creating GML by being presented the Gardels award.<ref>{{Cite web|url=http://portal.opengeospatial.org/files/?artifact_id=9477|title=award citation for Ron Lake}}</ref> The citation on the award reads “In particular, this award recognizes your great achievement in creating the Geography Markup Language, (GML), and your uniquely sensitive and effective work to promote the reconciliation of national differences to promote meaningful standardization of GML on a global level.” Simon Cox (CSIRO)<ref>{{Cite web|url=http://portal.opengeospatial.org/files/?artifact_id=16038|title=award citation for Simon Cox}}</ref> and Clemens Portele (Interactive Instruments)<ref>{{Cite web|url=http://portal.opengeospatial.org/files/?artifact_id=22940|title=award citation for Clemens Portele}}</ref> also subsequently received the Gardels award, in part for their contributions to GML.