Just about a week ago, I finished my summer semester for my work at #Eastern Illinois University. This has been a regular occurrence for the last few years. However, this summer term was my last and so it represents a milestone. The course was actually an independent study in preparation for the Fall term internship. To be truthful, the label of ‘internship’ may be a little misleading. I am working on a semester-long project that will end in a paper, presentation and product. My Employer is listed as the sponsor, but I was not asked to do the project nor will the end product will not be used by Cloudpoint Geospatial.

At any rate, I am very excited about the project.

The core of the project is nothing new; just a GIS professional looking into industry Open Source products or FOSS4G. The twist is the pivot. I am a literal certified expert in Esri’s ArcGIS Enterprise product platform and stack. I wanted to explore this space and see if there is any alternative to Esri software in the enterprise GIS ecosphere. Furthermore, can such an expert easily make a transition to an open source product, thus proving that the knowledge gained over 19 years is based on industry standards and not simply knowing a piece of software.

Simply put, have I wasted my time?
(That kind of personal reflection is written here in an unread blog and not in the final document)

I narrowed the definition of “Enterprise GIS” down to 5 components. Not based on Esri, not based on FOSS4G. Rather, based on the industry and what I have seen it become.

Enterprise GIS
  • Relational Database Management
  • Spatial Database Management
  • Local Desktop Editing & Visualization
  • Web-Publishing
  • Thin Client Editing & Visualization

Now, I will argue for these five, maybe even six, but no less. The year is 2020, not 1995. The ability for multiple people to edit a RDBMS from their desktop is no longer the totality of what is considered an “Enterprise Application System”. Conversely, I would argue that SaaS is not entirely enterprise as well. I know too many local governments and rural midwestern areas to rely completely on web-hosted applications.

Here’s what I looked at

MySQL, SqLite, PostgreSQL & SQL Server

Spatial Database:
ArcSDE, PostGIS, SpatiaLite, TerraLib

Desktop Editing/Viz:
ArcGIS Desktop (any Flavor), QGIS, gvSIG, GeoMedia

Web Publishing:
ArcGIS Server, GeoServer, MapServer, Map Guide

Browser Editing/Viz:
ArcGIS Portal, OpenLayers, Qfield, Leaflet, gvSIG Mobile

I assessed all of these packages on several criteria including my own personal comfort level. Remember; the purpose was to allow space for me to convert, not what leads the FOSS4G community. I may ultimately wrong in my assessment but in the end, this upcoming fall semester I will use PostgreSQL -> PostGIS -> QGIS -> Geoserver -> OpenLayers to create an Enterprise Open Source GIS stack on an Amazon AWS instance.

Disclaimer; the EC2 instance will be on a 2019 Windows Server AMI. (go ahead & hate, but I don’t have time to learn Unix bash commands)

Here is the Paper. I will post the result of the Fall semester in December.