A couple of my Spatial Networks colleagues and I spent most of last week in St. Louis, Missouri at the FOSS4G North America conference. Following up on attending the international FOSS4G conference in Boston, this is the most time I’ve spent with this community in quite a while. There has been no particular intent to stay away, but the ebbs and flows of life can result in absences from the scene.
I’m always struck by the diversity of imagination and content at every FOSS4G event and this one was no different. I am relative late-comer to FOSS4G events, having not attended one until the 2011 conference in Denver. That was the year of what I call ‘yet another javascript library’ because it seemed there were a lot of geo-centric libraries presented that year. It was also when I first saw CartoDB. The content, from my viewpoint, was very tech-focused, due primarily to a lot of new tech arriving on the scene and shaking up how even the more forward-leaning FOSS4G world was thinking about doing things.
Fast forward to St. Louis in 2018 and we’re on the back side of a couple of tech shakeouts, with some of the newcomers from 2011 now established veterans. Not only have we centered on just a couple of Javascript libraries, but R, not discussed in any session I attended in 2011, is now a well-established geospatial tool that is almost passé at this point.
The two main constants between 2011 and 2018 are PostGIS and GDAL. GDAL remains the workhorse it’s always been and was the beneficiary of a successful ‘barn raising‘ campaign to fund needed improvements. I’ll share some thoughts on that in a future post.
PostGIS and PostgreSQL have, of course, outlasted most of the NoSQL pretenders to remain the primary spatial data management platform in FOSS4G stacks (and maybe a few commercial stacks as well).
When the viability of the technology is a given, what’s is a technology conference to focus on? That answer, based on FOSS4G-NA 2018, is productivity, which places focus squarely on people instead of technology. That’s really as it should be.
In many cases, the ‘people’ that talks were focused on were ‘developers.’ Organizations such as Crunchy Data and LocationTech are focused on issues surrounding the deployability of systems based on FOSS4G technology. Notably in the age of GDPR and high-profile data breaches, those two organizations, along with Boundless, are placing a high-priority on validated, audited, secure builds of tools familiar in FOSS4G stacks. Security needs to be baked in from the beginning, so these organizations are providing a valuable service that allows developers to focus on solving business problems.
Even in more technical talks, such as one on OpenLayers 5, developer workflows were baked into the talk. Additionally, there was a strong theme of devops and containerization using Docker and Kubernetes across many talks. Gone (or perhaps hidden), for the most part, are Rube Goldberg means of deploying and maintaining systems built on FOSS4G technology.
For a number of reasons, I think open-source geospatial technologies had higher hurdles to jump to gain mainstream acceptance in the US market. My hope is that evidence such as the list of companies supporting the GDAL barn raising means that we are past that and that events such as FOSS4G-NA can continue to focus on the many people-centric reasons this technology is built and licensed the way it is. Last week in St. Louis felt like that turn has been made.
