Looking back at 2019 Geospatial Python
It was a busy year in 2019, except for a short holiday in July.
2019 came and went. Lets begin with how my tech stack changed in 2019.
Technology Stack as of December 2019
The basic core has not changed in the last several years, including Python, PostGIS, Postgresql. What is new this year is Google Cloud. Google Cloud Storage and Compute Engine were front and center this year for my new applications and deployments. The experience was smooth and polished in comparison to the AWS world which is also getting better I may add. Azure is looking good aswell, and I am looking forward to seeing what happens in 2020 within these big 3 players.
GDAL in 2019
Installation of geospatial libraries was still a pain in 2019. Docker of course makes this pain less painfull but nonetheless I do not feel comfortable in saying that getting GDAL to jive with your application stack works out of the box. I was hoping this would be a solved problem but if you scour the internet there are still more than a few individuals posting problems with GDAL. The problem is hiding not so much in the installation of GDAL but in the way in which other applications try to find the GDAL libraries.
Postgresql / PostGIS 2019
Serving vector tiles directly from a Postgresql with PostGIS extension is all the rage in 2019. I believe in 2020 this will continue and finally eliminate the need for any geoserver like server (sorry Geoserver) to render your vector tiles.
The desktop GIS world is alive as ever. QGIS has become the defacto GIS Desktop of choice for me over the last 5 years. ArcGIS I know you since 1999 the power, the feature count, the usability, the stability, the market penetration you are the king of GIS. QGIS has what I believe covers 95% of your needs.
QGIS has a very bright future and I would like to give a big shout out to the entire QGIS team for building such a great application!
Deployments in 2019
Gitlab, Gitlab CI /CD and Docker, Docker-Compose have taken over my deployments.