If you ever want to test a potential hire for a GIS position give them a task to accomplish using a different GIS system than that which they are familiar with. Having taught GIS for almost 10 years, it’s clear that those students who are capable simply read the documentation or do a search online for the best way of accomplishing their desired task. So presenting a potential recruit with a new GIS system will quickly reveal whether they are familiar with the concept of figuring things out on their own by searching online. Once you are familiar with the GIS concepts, any GISystem should be straight forward.
I have been using ArcGIS for too long, its saving grace for me, over the last four years, has been the arcpy python module. I do also still teach using ArcGIS but that’s something I’m trying to change. The arcpy module involves dealing with less of ArcGIS’s inconsistencies and frustrations. It still has glaring cartographic limitations and a complete lack of meaningful support besides other users.
The price of the ArcGIS is obscene and esri’s business strategies are exploitative. For all this money you get crippleware, software that has obvious features limited or removed so that you buy the more expensive product level. If you want to do an intersect or union of more than two feature classes then get ready to pay a lot more. Esri is established, largely has a monopoly and clearly behaves as such. They offer poor service, a buggy slow product with maddening DRM. Esri scares those looking at alternative GIS systems using the following policy. If you do not pay your GIS fees for one year (say 2015), as you wished to explore alternatives, and for some reason wish to return to the cult of ArcGIS the following year (2016), you will have to pay for the year you did not subscribe for ArcGIS (2015) in order to renew your licence. All these points explain how the president and founder of esri, Jack Dangermond, is about the 700 richest person in the world.
I have been using R for my automated cartographic tasks rather than esri’s arcpy for a few years. A few weeks ago after wanting to perform spatial analysis for personal tasks I decided to give QGIS a try to determine the catchment zones of schools in the town of Beverley, UK. There is no map available of something that is clearly spatial. The only data available is a pdf file listing which streets are associated with which primary and secondary schools. This seems like an easy join away from a map. One of the most challenging steps was finding a decent program that converts PDF tables into a tabular text file. This complete, the next few operations were extremely quick and satisfying.
The PDF contained multiple names, formats for the same schools. QGIS classified them differently. Using Google refine quickly fixes this:
Overall using QGIS for the first time was extremely straight forward, fast and efficient. I much prefer how QGIS deals with projections and exporting data compared to ArcGIS. It appears QGIS considers usability. I didn’t have any lags or slow load times that I associate with ArcGIS. I did have it crash on me once – something I associate with ArcGIS versions < 10. It was a good experience, I have installed it at the office and hope to be able to teach using QGIS in the future.