It’s rather fitting that the second plenary talk on Wednesday had to do with “firehose” applications since the FOSS4G North America (FOSS4GNA) conference was something of a firehose in itself. Despite the fact that the event was smaller than the worldwide event in Denver back in October, I came away with the same “full brain” feeling.
Of course, given the recent production release of PostGIS 2.0, that was kind of the big story for this event. I attended a number of PostGIS/PostgreSQL-related sessions and came away with lots of new information. I especially enjoyed Paul Ramsey’s “what’s new” talk on Wednesday. One thing I enjoy about his talks (here and in Denver) is that he’s not afraid to throw sample SQL up on the screen. It’s one thing to hear about a new feature but it’s another thing entirely to see a concrete example. Some may find the idea of raw SQL in a presentation abhorrent but it worked for me.
I’ve already started playing with PostGIS 2.0 but haven’t gotten terribly far so these sessions helped a lot. It’s already been pretty well vetted by the community during it’s various pre-release stages so it seems remarkably stable. With the addition of topology and raster analysis in addition to 3D/4D indexing and new vector functions, PostGIS probably handles the majority of what most users would need for GIS analysis at the database level. I’m excited to dig deeper.
I also attended a few web-mapping themed talks. If there is one word that consistently and repeatedly passed the lips of most of the presenters, it was “GeoJSON.” I’ve blogged about GeoJSON in the past but it really is the oil that makes the modern geo-web run smoothly. The rate of adoption it is receiving rivals only one other de facto standard that I can recall: you may remember something called the shapefile. The fact is that anyone who’s been hunting through their file system looking for the “next shapefile” has been looking in the wrong place. It lives behind that nice map in your browser. The shapefile took off because it solved a problem for what was the frontier of GIS in the 1990s; the transition from heavy workstations to desktop mapping. GeoJSON solves the problem of delivering vector spatial data for today’s frontier; the transition from the desktop GIS to web mapping. It’s that simple. We don’t need another file format to solve an old problem that already has a solution.
Of course, there was a lot more. I got introduced to CartoDB in Denver, have experimented with it since then and was able to get an update on its 1.0 release at one of the sessions. It remains a compelling option for spatial data hosting and Vizzuality are actively engaging with the open-source communities that build the tools they use. On the cartography side of things, Stamen continues to show how beautiful maps can be made with open-source tools. Similarly, MapBox showed how they efficiently deliver high-quality maps on the web. For my money, the unsung hero of what MapBox is doing is UTFGrid. I hadn’t delved into it deeply until I saw the announcement that OpenLayers now supports it. UTFGrid is breathtakingly elegant in its approach to delivering richer content with tiles. I expect that, in a few years, UTFGrid will be a standard part under the hood of any web mapping experience, like tiles and slippy maps are today.
The only consistent criticism I heard was that most of the sessions were “too technical.” I can see some validity to this, given that the event was held in Washington, DC and many higher-level government or policy types who may have been seeking to conceptually understand open-source geospatial may have been put off. Commercial vendors have addressed this problem with the establishment of “executive tracks” at their events. That said, if that’s the biggest criticism that can be leveled, then FOSS4GNA was a rousing success in my book.
I have resolved that I will take at least one of my kids, if not both, to the next FOSS4G event that I attend. The attitude of the open-source crowd is one that I want to rub off on them. Aside from the energy that is exuded from a gathering such as this, the open-source world is full of people who see a problem and just set out to solve it. I want my kids to be exposed to that kind of initiative and understand the power that can come from harnessing it, even if they never go into a technical field.
My understanding is that FOSS4GNA came together very quickly. Given that, it was a very high-quality conference and the organizers should be commended. Early tongue-in-cheek commentary referred to it as the “OpenGeo User Conference.” I felt like there was good balance and, when I attended sessions given by OpenGeo staff, I didn’t feel like I was beaten over the head with the OpenGeo brand. All-in-all, I found my day-and-a-half at FOSS4GNA to be very beneficial and I’m hoping it becomes an annual event. (Update: It looks like next year’s FOSS4GNA will be in Minnesota.)