There were, of course, many different topics presented. Yet the growing complexity of FME projects and systems was a common theme I noticed threading through much of the content. The issues caused by, and the techniques for dealing with this complexity bear a strong resemblance to those that software developers have already been wrestling with for decades. This is no accident since, among other things, FME is a graphical programming language with a pipe-and-filter structure for arbitrarily complex data and geodata processing. FME practitioners are only now discovering the value of disciplined software architecture, good logic structure/representation, agile project management, automated testing, continuous integration, proper version management, structured documentation extraction, etc. The platform itself hinders some of these methods, especially concurrent version control for teams, but I have made the great folks at Safe Software aware of the scope of these limitations both before and during the conference. They were very open, gracious, and clearly motivated by the challenges and potential solutions we discussed. I’m sure we will be seeing some dramatic improvements in this direction n the coming years.
Other great topics included performance tuning and optimization, geometric tolerance factors, indoor mapping, using FME as the backbone of a large enterprise’s integrated data architecture, and many other worthy topics. Additionally, Dale, Don, and various other Safers provided us with a sneak peek of the future of FME, including the continuing build-out of FME Cloud capabilities, new transformers and other components in the core product, and the architecturally sound migration of the platform to a Docker container model.
In closing, I’d like to paraphrase a great line from one of the plenary/keynote addresses: In the modern world, data has become a more valuable resource than oil. This is just as true for geodata as for traditional data.