The Free and Open Source Software for Geospatial North America (FOSS4GNA) conference was held in Washington, D.C. on April 10-13. Hot on the heels of FOSS4G in Denver last fall and a warm-up for FOSS4G 2012 in Beijing, China, the event drew 360 attendees, with many coming from the local area.
More coverage of the event can found at All Points Blog.
The first half of the opening plenary was an introduction to the history of open source and the open source geospatial software stack. I wondered if that was still needed. Per Paul Ramsey, conference chair, of OpenGeo, who gave that intro, it is. While many in the audience are longtime users of open source, he estimates some 50% are still new to the concept, or new to the geospatial offerings. I ran into several people who fit that description. One fellow works as a geospatial analyst for a major geospatial player. He uses proprietary geospatial software daily and was sent to the event by his boss.
This event is really a developer conference. The technical discussions seemed to hold the audience’s rapt attention. My sense was the less technical presentations (project-based ones, the app-focused ones, the “we did this, then this, then this...”) drove a good percentage of attendees to their laptops to surf and check e-mail.
Business models and how to make money with open source were notoriously missing from the event, save one small mention in the opening keynote and a second in the “gaps” panel. I’m not sure if that’s because this subset of the community (mostly technical as I observed above) is not the one worried about such topics or if there’s a sense of calm about providing services, rather than products, as a business model.
How the Map Looks Matters!
There was a surprising amount of discussion on app and map design and layout. The idea that both the user interface and the maps need to be attractive and responsive popped up in variety of sessions. “Ugly” and “hard to use” are no longer acceptable to the open source community.
iPads Nowhere to be Seen
I did not see a single iPad in two days. My explanation? This is a technical group; they don’t use iPads. They do, however, use Linux-based netbooks.
Big Companies are Quiet
Both Esri and Google were sponsors but neither spoke in a paper session, nor made much noise at all. Smaller companies whose staffers spoke (MapBox, CamptoCamp, OpenGeo and many others) were respectful and calm when addressing those companies’ open source and proprietary offerings.
There’s an elegant dance going on, one that I believe will ultimately lead to many, many hybrid (open source and proprietary implementation) solutions.
Autodesk, which is responsible in so many ways for the creation of the Open Source Geospatial Foundation (OSGeo), was nowhere to be seen. I heard no mention of MapGuide Open Source, which is an OSGeo project. While its e-mail list has regular activity, it does not seem to have the widespread use of other OSGeo projects such as PostGIS and OpenLayers.
Several attendees’ and sponsors’ organizations were looking for staff. However, I didn’t meet anyone who was out of a job. I did see a tweet from an intern looking for some input from a sponsor. Is this because this self-select group was sent by their employers? Or, is demand for this select set of geospatial technologists so very high? Or, is this just a hot market?
Massimo Di Stefano, who attended the conference as part of his job at RPI, has a great “job” story. His work with the first OSGeo local chapter in Italy, while in school, got him known in the community. That led to a short-term job at Woods Hole Oceanographic Institute and now a three-year position on an NSF grant-funded project at RPI.
As Sean Myers (@seanmyers) of NBT Solutions noted in a tweet: “Lots of random conversations at #FOSS4GNA about the demand for technical talent. Where are all the spatial db admins and programmers?”
A Real-time Game-like Modeling System for the World
I heard a description of the same system vision from three different people from three different organizations in a single day. The tool would allow a variety of data inputs, and model, on a map, “what if” scenarios built on various parameters (perhaps set by sliders) with results rendered in real-time. As an example, the predicted population distribution map for 2250 would change in real-time as the sliders setting rainfall, disease and carbon dioxide levels were adjusted.
Who was talking about that? For one, Robert Cheetham of Azavea. He spoke of those sorts of systems as key users of his company’s new GeoTrellis (open source) project, which aims to provide that sort of performance for geospatial modeling. And then there was Schuyler Earle, who described such a system in the “wish list” as part of the “gaps” panel. And Scooter Wadsworth of Sanborn Map Company, who described some experiments to create that sort of real-time modeling using raster data with processing via MongoDB, a NoSQL database.
The open source geospatial community continues to attract new developers and users. The challenge to conference organizers going forward is to provide offerings for both parts of this vibrant community.