The Future of GIS, a Study of Conferences — Part Two
So FOSS4G released their sessions on May 2nd, when I was in Boston at for another conference. (which I have a bunch of blog entries lined up to write on) Since returning to the mountains I’ve been playing catch up with work and haven’t really had time to dedicate to the text analysis until now. I guess this is just a long way of saying “off my back. I have a life.” Regardless, let’s take a look at how FOSS4G sessions break down.
Methodology:
FOSS4G breaks down its presentations into 5 categories:
- Business
- Education
- FOSS4G Development
- Government Policy
- New Technology
- Science
That being said, for the purpose of this analysis we are tossing out their classifiers. Unlike the ESRI UC or Developer’s conference where the technological baseline is a given (ArcGIS Server, ArcSDE, EZ PASS Spatial Online), each session at FOSS4G could deal with completely different technologies. You could have a something classified as Business, and it could deal with selling POSTGIS in a Serverless Architecture. As a result of this, I tossed out these classifiers, I’ll do the same for the UC, as they also classify sessions by area of interest.
As a result of the “Non Standard” technological base, the text analysis for FOSS4G was a bit more detailed, with few more “if this than that” type of analysis. The most prevalent of this was Cesium, should it be Javascript or should it be under 3D? In the end, there are so many Javascript frameworks out there that if you chose Cesium, you’re OBVIOUSLY looking for 3D.
We’ll also go into the workshops (self-promotion I’m leading the Spatial Sql Workshop for Rookies on Monday Morning. I’m hilarious at 0800 in the morning after traveling across the country the day before), in another blog entry, as FOSS4G is a thick conference, with a bunch going on.
Lastly, I’m not including JS Geo. The reason for this is <spoilers>ITS ALL JAVASCRIPT </spoilers>
All things PY ended up in Python, and all things .JS ended up in Javascript.
Since FOSS4G isn’t as cut and dried as the Dev Summit, we’re going to go with straight word count. Only one count per session.
Va-va-va VOOM
Results:
Summary:
Wow, most of this isn’t too surprising to me, but 3D coming in first, I am bit stunned by this. I expected it in the top 5, but not to be mentioned more than QGIS. Removing the Cesium references it puts it down less than POSTGIS, which is where I figured it would be. This shows that there is a real demand for 3D, both desktop and on the web within the FOSS4G community.
In a two way tie for second place, we have QGIS, and Python. None of these are surprising, since QGIS is the defacto desktop client of the FOSS4G community and Python have all generated huge amounts of interest in the Spatial and FOSS4G communities for a number of years.
With Q and its use, and customization having a large number of sessions was pretty obvious. I’m sure with gvSIG winning open source awards in Europe, and with the advancement in spatial statistics within GeoDA, this may not be the case during future conferences.
The number of Python presentations also aren’t eye-popping. With it being the most popular language within software development, the real surprise is that it is not higher.
Tied for third place, we have PostGIS and Javascript. It should be noted that the reason I believe Javascript is in 3rd place is because of JSGeo occurring on Tuesday (Tickets are currently sold out for JSGeo. Another 30 tickets will be released in 17 July). Within just FOSS4G itself, the Javascript presentations are all around web viewers and a couple of Node.JS sessions.
PostGIS (caveat which I love)…
<narrator> He even made a t shirt to show his love….for a database
…has sessions that run the gambit from use, to extending its capabilities, to using pgRouting and how to return data in various data outputs. Must importantly, how to use PostGIS on how to create Microservices. Which is important because…well just keep reading.
For fourth place we have a three way tie, R, Geoserver and Spark. After the ESRI Dev Conference analysis having R so low, I am surprised to see R this high on the list. But with its recent popularity in the Data Science community and the cross pollination between Spatial and Data Science. R is on the rise across the board. I have a slotted upcoming blog post about the rise of R in spatial.
Like Q is to desktop, Geoserver is to defacto standard of Web Servers within the FOSS4G community. Pretty typical array of presentations, management, extension and use. It’s not as sexy as say Python or 3D, but it is part of the plumbing that makes it all work.
While I was surprised that 3D takes the top spot, I am bowled over that Spark is this high on the list. Even with categorizing GeoMesa with Spark, I didn’t expect it to be in the mid range. I guess people want their geo-analytics done quickly.
Rounding out the top five, and numbers 9 and 10 on the chart, we have Serverless and Docker. Serverless, for the purposes of this analysis was those sessions which allow you to serve data over the web without using GeoServer. Like the microservices references in the PostGIS sessions. Most of these sessions center around architecting these solutions. Serverless tech is directly related to…
Docker, which allows you to containerize your apps and data and you can present it in a serverless framework. With both Serverless and Docker both appearing on the list, and if they were combined into a single grouping (which maybe I should have done) it would be tied with Q and Python. Meaning almost twice as many as Geoserver. This is an area to keep a tab on as it progresses forward.
Honorable Mention:
There are two machine learning sessions, and if you’re in Spatial and think this is really advanced tech, it is not. In the Data Science community its machine learning is ubiquitous with their analysis. Not used by that guy in the back with the Masters in Statistics, but used by everyone when its needed. If you deal with data, at all, and are going to FOSS4G, hit one of these. If you can’t hit them, find me, and I’ll point you to some tutorials or something. If I’m not at the conference, I’ll be at the Tiki Bar with @bitnerd.
Conclusion:
FOSS4G is thick, but in a good way. Most of the results on this list are not shocking, but the advancement of both Serverless Tech and Spark are great. It puts us in closer step with the greater information technology field in general.
This is starting to get long, and as I said, FOSS4G is THICK this year. I’ll do workshops and a verbose rap up in another entry.
Second Verse of this blog can be found here
