Work stories: 1986

 Geo/SQL [Previous ¦ Next]

My start in geographic info systems (GIS) was at the Geological Survey of Canada: I compiled Arctic Islands geology in a database management system, and sent files to Ottawa headquarters for referencing against ESRI coverages (electronic map sheets) prepared at the federal mapping agency in Canada's capital. I went on to joint-venture with a geophysical company, who also needed such digital data capture services. I ended up working with Bill who later founded Geo/SQL, because we had both devised an AutoCAD front-end to Oracle and DB2, which was in high demand in the oil industry. I initially met people at Chevron, Gulf and Texaco in Calgary who saw an immediate need in posting wells and plotting pipeline routes (seismic was deemed to be a more difficult and later task). Bill was a surveyor who programmed a 10Kb geospatial kernel that handled re-projections of data in the AutoCAD window, and fetched data from Oracle or DB2 via structured query language (SQL).

Side story: So near yet so far: this a a pretty small world at the time in geo-computing. Bill and I took this idea of a geo kernel to spatialize Autodesk's AutoCAD by calling up then head John Walker. We got his PA at the time Carol Bartz - later to lead Autodesk and become famous at Yahoo! (exclamation mark part of the name brand) -   and various conversations lead Bill to travel to Sausalito CA then Autodesk HQ, but nothing came of it. Imagine our brush with fame, had AutoCAD been spatialized... not only would this web page look very different, but Autodesk might've given Intergraph and Esri a run for their money at the time! Enigma and  Munro give follow-ons to this story.

Neither SQL nor GIS meant much to anybody as this was still pretty innovative. Open was not yet a buzzword and Unix was actually platform-dependant, so we decided to work on PCs running DOS: it came without graphics however (Windows was yet to become the standard it is today), so AutoCAD was chosen for its rich graphical interface; its scripting language AutoLisp was also open unlike, say, Intergraph on workstations that remained proprietary . [1]My efforts faltered in both petroleum and mining, where the need for GIS wasn’t yet perceived in Western Canada (see Enigma above). Geo/SQL went on to a turbulent history, change of hands in Colorado then Japan, and finally back to Bill who contracts to this day in Texas.

1: Bentley eventually broke the binary code and started MicroStation on PCs. Binary formats were common then (s.a. ESRI coverages as well as Autodesk and Intergraph CAD files) because of hardware restrictions that demanded files be small, and large datasets broken into tiles for adequate processing. Being binary also helped maintain proprietary formats, in days when files were deemed to hold competitive advantages - open standards and file exchange had to wait a while (see Prizm).