Keynote
Mark Monmonier from Syracuse University offered the keynote titled "Customization and Control in Geospatial Technology: Unprecedented Leverage and Unintended Consequences." If you've read any of his books the underlying theme was familiar: Be careful with the technology! He offered a few scary stories on control, tracking, finding sites for low level radioactive waste in New York State and gerrymandering.
The presentation was only weakly connected to GIS, which was a nice change. (Monmonier didn't even mention Google!) Monmonier also noted the importance of the history of cartography/GIS for today's professionals. As I looked around the audience I wondered if today's cartography students still read his How to Lie with Maps in school as I did.
One comment Monmonier made highlighted the changes in the last decade. He noted how citizens of Allegheny County, NY, after having potential sites for low level radioactive waste located there, sued for access to the GIS database that was used in the siting process. They got it, but it "was not in a standard format, ArcInfo or ArcGIS" since the consulting firm who did the work, used its own GIS software. This was back in 1990s.
Sessions
A session on mobile, field and data storage offered no new ground on rugged/semi-rugged/non-rugged mobile hardware. We did get to hold all the different devices, however. The presentation on storage highlighted the value of virtualization. Virtualization? It's basically making a mix of storage devices look like just a few servers. It does two things: allows for the back ends to be changed without the users knowing and allows for a single point of entry for users. The final paper on communications highlighted a theme that ran throughout: decreasing complexity increases usability. Indeed.
A "birds of a feather" session on "credentials in GIS" yielded some basics on certificate programs (a tool to get a job), GISP (three folks had them and the general consensus among them was it was the future, though few in attendance seemed to know much about it), and online programs (those from schools offering bricks and mortar programs were skeptical). My big contribution to the discussion tied into one faculty member's note that you can and should apply for a job for which you are not qualified, especially if you note that you are willing to learn. I added that's its even better to show that you are both willing and able to do so. How? By noting how you've taught yourself something – a programming language, how to surf, etc.
The high point of the first day was a conversation between Kevin Flanders of PeopleGIS (our open source columnist) and Saul Farber of MassGIS. They showed off and commented on how open source, collaboration, standards and other factors are redefining GIS. They paused now and again to answer questions – which ran the gamut from issues of data licensing to the ability to share GPS breadcrumbs in Google Earth to the value of 3D. Both men felt strongly that intertwining of open source and the Internet is the key to what's going on, and will go on, in geospatial technology.
A data point I thought was valuable: Farber asked the audience (maybe 50 or 60) if support for open source was still something that would (or had) prevent(ed) them from selecting open source. One attendee simply said it was easier to use, in his case ArcIMS, out of the box. Another said open source was okay for "temporary" apps, but would need more robust support for what I'd term "real" uses. The general consensus, I thought, was "no, not really."
A romp through some open source built apps that looked (and acted) like Google Maps (ka-map based) helped open eyes, too. The listing of projects behind the demos they showed also prompted interest: MapServer, GeoServer, Community MapBuilder, MapBender… Those in the audience asked about PostGIS and GRASS.
A three paper session on CAD/GIS interoperability ran down the processes for sharing data between the two technologies. While both ESRI and Autodesk products have new features that improve sharing data, I still feel we've not come far enough on this topic, as the final paper suggested. There are more dialog boxes and tools, but you still need to know quite a lot to share data effectively.