Hacker News Reader

Comment by jld | Hacker News Reader

As a mid-career software engineer who would be interested in getting into the "geospatial industry",

A) What technologies should I learn to allow me to get my foot into the industry? I have historically focused on being a UX focused web engineer. I'm sure that experience could surely transfer, but am interested in learning the technologies and/or understanding that is more core to geospatial problems.

B) As someone who loves maps and imagery, but does not want to work for the military industrial complex, how do I avoid point #3. "In geo, you either die a hero or live long enough to make the majority of your revenue from defense and intelligence."

View on HN

A: Break it into two parts, vector and raster. For the former, focus on Mapbox and associated libraries/standards (VectorTiles, Leaflet, GeoJSON, etc.). JS is the name of the game in vector land—it’s amazing what you can do all on the front end now. For backend/DB tech, look at PostGIS.

For the latter, I’d focus on Python to start. Look at GDAL (and it’s various Python interfaces/wrappers like rasterio), Cloud Optimized Geotiffs, and the SpatioTemporal Asset Catalog spec.

In both cases, watch what the leading companies are putting out (Mapbox, Azavea where I work, Development Seed, etc.) and not what OGC is doing. One is a leading indicator, the other is a lagging one. Sorry OGC, I still love you and you play an important role.

B: Get comfortable with the idea of working with Defense. For better or worse, they’re the ones with the budget to take on humanitarian work, disaster response work, even climate change adaptation work. That’s my advice. Possible to avoid it altogether, but only if you work on an app that is domain-focused (eg Trulia) and incidentally geospatial.

B) Look into healthcare systems or public health. Defense holds sway because of its budget size, but it's not the only fat cat in town.

Epidemiologists are nice, but they're expensive, rare, and usually bad at GIS programming. Most governments will happily hire a GIS analyst. They can't get enough maps and data dashboards. And most of the ones they do have are terrible (misleading, slow, convoluted, or just plain not useful).

Reply on news.ycombinator.com