The move from proprietary software to open source in the geospatial industry is something I’ve been personally engaged with for more than 15 years. So I’m pleased to have recently led a session called “How and Why Open Source Software Has Become Mainstream for GIS” at the very fun and informative SDI-Open 2017 workshop in Washington, D.C. The workshop was a pre-conference event for the 28th International Cartographic Conference. The SDI-Open folks are essentially commissioners of the International Cartographic Association, and are the instigators (along with OSGEO) of the growing “Geo for All” education initiative.
The success of open source is a topic worth hours of discussion. Unfortunately, as we only had one morning at SDI-Open 2017 (and we had to share the time slot with a very excellent overview of GRASS GIS, Blender and data visualizations put on by Vaclav Petras, Anna Petrosova and Payam Tabrizian of the NC State Center for Geospatial Analysis), I went straight to the key points:
- Open source makes for better GIS software.
- Free, open source software gives users and developers more control and allows us to work in new, better and more effective ways.
- These benefits have moved open source into the mainstream of GIS use.
Open source makes for better software
Having worked for both proprietary (Ionic) and open source (OpenGeo/Boundless and DigitalGlobe | Radiant) geospatial software companies, I have seen firsthand how open source improves collaboration compared with proprietary software. In the proprietary world, software intellectual property (IP) is tightly controlled and only available to the developers the IP owner allows (usually company employees). This creates a pernicious situation where the company has a strictly limited developer roster, and developers cannot work on the software anywhere else—so developers and owners are both isolated and stuck with each other!
Open source’s inherent ability to improve collaboration in turn stimulates and accelerates innovation. The open level of interaction among developers and organizations means open source software is available to people with different points of view—diverse backgrounds, goals, perspectives, expertise—spurring creativity. It also means the openness of the software creates the conditions for a rich ecosystem for development. In my early proprietary software work, essentially all of our core developers worked in the same office. When I moved to OpenGeo, we had open source core developers in Austria, Italy, the Netherlands, Argentina, Australia, Canada and the U.S., and collaborators almost everywhere (well, not Antarctica). Talk about diverse perspectives and a rich ecosystem!
Another key driver of innovation is the sustainability of open source—good ideas (in the form of software code) aren’t lost. Open source code is always there to be built on by any set of users, even across programs, projects and organizations. Innovation is easier when you don’t have to start from scratch, but can build software infrastructure on successful bits begun in other programs or endeavors.
Open source gives users and developers more control
As someone who has produced and sold both proprietary and open source software, I have seen how this shift of control with open source changes everything. Open source takes control of the software roadmap and feature set away from the proprietary software owner and puts it into the hands of the users and producers of the software. Now the users and developers can decide on timing and speed of release cycles, and users can decide on such things as level of effort, partners and developers to work with, and the costs they are willing to bear to achieve certain features, e.g., maintaining the software.
It is commonly noted in free open source software (FOSS) circles that it’s “free as in freedom,” not “free as in free beer” (see this eloquent perspective from my colleague, Ilya Rosenfeld). I agree wholeheartedly, but I also believe the huge problem of rapidly rising software license costs was a primary push driving the switch to open source. This switch initially occurred with the need to scale web services and accelerated with the move to the cloud. Today, with SaaS and PaaS obscuring license issues somewhat, the control open source gives users and developers (e.g., roadmap control, options for maintenance, etc.) pulls organizations toward FOSS.
Open source and GIS
I could go on and on about how collaboration, innovation and control advantages have led to the mainstreaming of open source GIS. I could also point you to the growing roster of organizations supporting and using open source GIS. But I think the most useful discussion is how we at DigitalGlobe are building and use open source software for a huge range of geospatial applications every day.
For big data, we’ve led the development of GeoWave to leverage the scalability of a distributed key-value store for effective storage, retrieval and analysis of massive geospatial datasets. GeoWave enables enterprises to map and query geospatial information at scales far beyond standard geodatabases. Our heritage is pioneering open source image processing, storage and retrieval with OSSIM and OMAR. We develop geospatial libraries and applications that are used in the cloud to process imagery, maps, terrain and vector data. On the cutting edge, we are using distributed cloud computing and our updated imagery repositories to automate feature extraction for the world’s coastlines (Beachfront and Piazza), developing raster analytics at large scale (MrGeo), and leading the way on high-volume data conflation (Hootenanny).
I urge you to attend FOSS4G 2017 to meet some of DigitalGlobe open source specialists, and to learn more about our software and the other great open source software tools that are in wide use around the world.