I watch this and can’t help but wonder: haven’t we been here before?
If we replace the phrase “neogeography” with the phrase “desktop mapping” then I feel like we had this conversation 10 (or more) years ago. The ironic part is that MapInfo was in the position then that neo-geo is in now.
Many GIS hard-liners argued that non-topological mapping packages could not be called GIS. “Real GIS” understood topology and could do analysis based upon topology. That argument was mostly forwarded by the Arc/Info crowd (although I don’t think ESRI was directly involved, it arose more from Arc/Info adherents). Software such as MapInfo, Maptitude, etc. was not “real GIS” and, therefore, their outputs could not be trusted.
Well, we see what happened: that market proliferated. ArcView 3.x became the most widely used GIS package around at the time. Other desktop systems (MapInfo, Manifold, etc.) beefed up their capabilities and ESRI eventually produced ArcMap. Anyone care to have that conversation again? I didn’t think so.
So, is neogeography GIS? I don’t know. What’s the measure? Those put forth by Mike Hickey are different than those put forth as arguments against desktop mapping years ago. So, really, the hurdle has just been moved. Maybe neo-geo isn’t GIS…yet. But does anyone really think the technology won’t advance? Take a look at FeatureServer. Take a look at the various spatial SQL implementations in PostGIS, SQL Server 2008, Oracle and many others. People are trying to solve that problem. If ESRI ever puts out their REST API, the neo-geo community will have the full power (maybe) of ArcGIS Server at their disposal. I have an idea: if MapInfo feels that neo-geo isn’t really GIS, why don’t they create a product line targeted at giving the neo-geo community “real GIS” tools. Whatever those are. Looks like a market opportunity to me.
Making decisions about the viability/validity of neo-geo based on the current state of the technology is a little short-sighted to say the least.